🕷️ Crawler Inspector

URL Lookup

Direct Parameter Lookup

Raw Queries and Responses

1. Shard Calculation

Query:

Response:

Calculated Shard: 168 (from laksa198)

2. Crawled Status Check

Query:

curl -X POST \
  'http://laksa168.int.ahrefs:8124/' \
  -H 'Content-Type: text/plain' \
  -H 'X-ClickHouse-Database: crawler3' \
  -H 'Authorization: Basic YXBpOg==' \
  -d 'SELECT getAhrefsURLFromUnparsed(src_unparsed) AS found_url, ifNull(toUnixTimestamp(download_stamp), 0) AS crawl_time, ifNull(toUnixTimestamp(props_url_first_seen), 0) AS first_indexed_time, download_http_code AS http_code, src_unparsed AS src_unparsed, src_root_hash AS src_root_hash, history_drop_reason AS history_drop_reason, meta_title AS meta_title, meta_descriptions AS meta_descriptions, attrs_boilerpipe_text AS attrs_boilerpipe_text, attrs_markdown AS attrs_markdown, attrs_readable_markdown AS attrs_readable_markdown, meta_canonical AS meta_canonical FROM crawler3.page_info_local FINAL PREWHERE (src_root_hash, src_unparsed) IN ((getAhrefsRootHashFromUnparsed(getAhrefsUnparsedNoserviceFromURL(\'https://www.tensorflow.org/install/pip\')), getAhrefsUnparsedNoserviceFromURL(\'https://www.tensorflow.org/install/pip\'))) FORMAT JSONEachRow'

Response:

{"found_url":"https:\/\/www.tensorflow.org\/install\/pip","crawl_time":1775799242,"first_indexed_time":1537650112,"http_code":200,"src_unparsed":"org,tensorflow!www,\/install\/pip s443","src_root_hash":"12537842311192732768","history_drop_reason":null,"meta_title":"Install TensorFlow with pip","meta_descriptions":[],"attrs_boilerpipe_text":"Install Tensor\nFlow with pip\nStay organized with collections\nSave and categorize content based on your preferences.\nOn this page\nHardware requirements\nSystem requirements\nSoftware requirements\nStep-by-step instructions\nPackage location\nPython version support\nThis guide is for the latest stable version of TensorFlow. For the\npreview build\n(nightly)\n, use the pip package named\ntf-nightly\n. Refer to\nthese tables\nfor\nolder TensorFlow version requirements. For the CPU-only build, use the pip\npackage named\ntensorflow-cpu\n.\nHere are the quick versions of the install commands. Scroll down for the\nstep-by-step instructions.\npython3\n-m\npip\ninstall\n'tensorflow[and-cuda]'\n# Verify the installation:\npython3\n-c\n\"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n# There is currently no official GPU support for MacOS.\npython3\n-m\npip\ninstall\ntensorflow\n# Verify the installation:\npython3\n-c\n\"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\nconda\ninstall\n-c\nconda-forge\ncudatoolkit\n=\n11\n.2\ncudnn\n=\n8\n.1.0\n# Anything above 2.10 is not supported on the GPU on Windows Native\npython\n-m\npip\ninstall\n\"tensorflow<2.11\"\n# Verify the installation:\npython\n-c\n\"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\npython3\n-m\npip\ninstall\ntensorflow\n[\nand-cuda\n]\n# Verify the installation:\npython3\n-c\n\"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\npython3\n-m\npip\ninstall\ntensorflow\n# Verify the installation:\npython3\n-c\n\"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\npython3\n-m\npip\ninstall\ntf-nightly\n# Verify the installation:\npython3\n-c\n\"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\nHardware requirements\nThe following GPU-enabled devices are supported:\nNVIDIA® GPU card with CUDA® architectures 3.5, 5.0, 6.0, 7.0, 7.5, 8.0 and\nhigher. See the list of\nCUDA®-enabled GPU cards\n.\nFor GPUs with unsupported CUDA® architectures, or to avoid JIT compilation\nfrom PTX, or to use different versions of the NVIDIA® libraries, see the\nLinux build from source\nguide.\nPackages do not contain PTX code except for the latest supported CUDA®\narchitecture; therefore, TensorFlow fails to load on older GPUs when\nCUDA_FORCE_PTX_JIT=1\nis set. (See\nApplication Compatibility\nfor details.)\nSystem requirements\nUbuntu 16.04 or higher (64-bit)\nmacOS 12.0 (Monterey) or higher (64-bit)\n(no GPU support)\nWindows Native - Windows 7 or higher (64-bit)\n(no GPU support after TF 2.10)\nWindows WSL2 - Windows 10 19044 or higher (64-bit)\nSoftware requirements\nPython 3.9–3.12\npip version 19.0 or higher for Linux (requires\nmanylinux2014\nsupport) and\nWindows. pip version 20.3 or higher for macOS.\nWindows Native Requires\nMicrosoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019\nThe following NVIDIA® software are only required for GPU support.\nNVIDIA® GPU drivers\n>= 525.60.13 for Linux\n>= 528.33 for WSL on Windows\nCUDA® Toolkit 12.3\n.\ncuDNN SDK 8.9.7\n.\n(Optional)\nTensorRT\nto improve latency and throughput for inference.\nStep-by-step instructions\n1.\nSystem requirements\nUbuntu 16.04 or higher (64-bit)\nTensorFlow only officially supports Ubuntu. However, the following\ninstructions may also work for other Linux distros.\n2.\nGPU setup\nYou can skip this section if you only run TensorFlow on the CPU.\nInstall the\nNVIDIA GPU driver\nif you have not. You can use the following command to verify it is\ninstalled.\nnvidia-smi\n3.\nCreate a virtual environment with\nvenv\nThe venv module is part of Python’s standard library and is the officially recommended way to create virtual environments.\nNavigate to your desired virtual environments directory and create a new venv environment named\ntf\nwith the following command.\npython3\n-m\nvenv\ntf\nYou can activate it with the following command.\nsource\ntf\/bin\/activate\nMake sure that the virtual environment is activated for the rest of the installation.\n4.\nInstall Tensor\nFlow\nTensorFlow requires a recent version of pip, so upgrade your pip\ninstallation to be sure you're running the latest version.\npip\ninstall\n--upgrade\npip\nThen, install TensorFlow with pip.\n# For GPU users\npip\ninstall\ntensorflow\n[\nand-cuda\n]\n# For CPU users\npip\ninstall\ntensorflow\n6.\nVerify the installation\nVerify the CPU setup:\npython3\n-c\n\"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\nIf a tensor is returned, you've installed TensorFlow successfully.\nVerify the GPU setup:\npython3\n-c\n\"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\nIf a list of GPU devices is returned, you've installed TensorFlow\nsuccessfully.\nIf not continue to the next step\n.\n6.\n[GPU only] Virtual environment configuration\nIf the GPU test in the last section was unsuccessful, the most likely cause is that components aren't being detected,\nand\/or conflict with the existing system CUDA installation. So you need to add some symbolic links to fix this.\nCreate symbolic links to NVIDIA shared libraries:\npushd\n$(\ndirname\n$(\npython\n-c\n'print(__import__(\"tensorflow\").__file__)'\n))\nln\n-svf\n..\/nvidia\/*\/lib\/*.so*\n.\npopd\nCreate a symbolic link to ptxas:\nln\n-sf\n$(\nfind\n$(\ndirname\n$(\ndirname\n$(\npython\n-c\n\"import nvidia.cuda_nvcc;\nprint(nvidia.cuda_nvcc.__file__)\"\n))\n\/*\/bin\/\n)\n-name\nptxas\n-print\n-quit\n)\n$VIRTUAL_ENV\n\/bin\/ptxas\nVerify the GPU setup:\npython3\n-c\n\"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n1. System requirements\nmacOS 10.12.6 (Sierra) or higher (64-bit)\nCurrently there is no official GPU support for running TensorFlow on\nMacOS. The following instructions are for running on CPU.\n2. Check Python version\nCheck if your Python environment is already configured:\npython3\n--version\npython3\n-m\npip\n--version\n3. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip\ninstallation to be sure you're running the latest version.\npip\ninstall\n--upgrade\npip\nThen, install TensorFlow with pip.\npip\ninstall\ntensorflow\n4. Verify the installation\npython3\n-c\n\"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\nIf a tensor is returned, you've installed TensorFlow successfully.\n1. System requirements\nWindows 7 or higher (64-bit)\n2. Install Microsoft Visual C++ Redistributable\nInstall the\nMicrosoft Visual C++ Redistributable for Visual Studio 2015,\n2017, and 2019\n. Starting with the TensorFlow 2.1.0 version, the\nmsvcp140_1.dll\nfile is required from this package (which may not be\nprovided from older redistributable packages). The redistributable comes\nwith\nVisual Studio 2019\nbut can be installed separately:\nGo to the\nMicrosoft Visual C++ downloads\n.\nScroll down the page to the\nVisual Studio 2015, 2017 and 2019\nsection.\nDownload and install the\nMicrosoft Visual C++ Redistributable for\nVisual Studio 2015, 2017 and 2019\nfor your platform.\nMake sure\nlong paths are enabled\non Windows.\n3. Install Miniconda\nMiniconda\nis the recommended approach for installing TensorFlow with GPU support.\nIt creates a separate environment to avoid changing any installed\nsoftware in your system. This is also the easiest way to install the\nrequired software especially for the GPU setup.\nDownload the\nMiniconda Windows Installer\n.\nDouble-click the downloaded file and follow the instructions on the screen.\n4. Create a conda environment\nCreate a new conda environment named\ntf\nwith the following command.\nconda\ncreate\n--name\ntf\npython\n=\n3\n.9\nYou can deactivate and activate it with the following commands.\nconda\ndeactivate\nconda\nactivate\ntf\nMake sure it is activated for the rest of the installation.\n5. GPU setup\nYou can skip this section if you only run TensorFlow on CPU.\nFirst install\nNVIDIA GPU driver\nif you have not.\nThen install the CUDA, cuDNN with conda.\nconda\ninstall\n-c\nconda-forge\ncudatoolkit\n=\n11\n.2\ncudnn\n=\n8\n.1.0\n6. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip\ninstallation to be sure you're running the latest version.\npip\ninstall\n--upgrade\npip\nThen, install TensorFlow with pip.\n# Anything above 2.10 is not supported on the GPU on Windows Native\npip\ninstall\n\"tensorflow<2.11\"\n7. Verify the installation\nVerify the CPU setup:\npython\n-c\n\"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\nIf a tensor is returned, you've installed TensorFlow successfully.\nVerify the GPU setup:\npython\n-c\n\"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\nIf a list of GPU devices is returned, you've installed TensorFlow\nsuccessfully.\n1. System requirements\nWindows 10 19044 or higher (64-bit). This corresponds to Windows 10\nversion 21H2, the November 2021 update.\nSee the following documents to:\nDownload the latest Windows 10 update\n.\nInstall WSL2\nSetup NVIDIA® GPU support in WSL2\n2. GPU setup\nYou can skip this section if you only run TensorFlow on the CPU.\nInstall the\nNVIDIA GPU driver\nif you have not. You can use the following command to verify it is\ninstalled.\nnvidia-smi\n3. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip\ninstallation to be sure you're running the latest version.\npip\ninstall\n--upgrade\npip\nThen, install TensorFlow with pip.\n# For GPU users\npip\ninstall\ntensorflow\n[\nand-cuda\n]\n# For CPU users\npip\ninstall\ntensorflow\n4. Verify the installation\nVerify the CPU setup:\npython3\n-c\n\"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\nIf a tensor is returned, you've installed TensorFlow successfully.\nVerify the GPU setup:\npython3\n-c\n\"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\nIf a list of GPU devices is returned, you've installed TensorFlow\nsuccessfully.\nPackage location\nA few installation mechanisms require the URL of the TensorFlow Python package.\nThe value you specify depends on your Python version.\nPython version support\nVersion\nURL\nLinux x86\nPython 3.10 GPU support\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl\nPython 3.10 CPU-only\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow_cpu-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl\nPython 3.11 GPU support\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl\nPython 3.11 CPU-only\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow_cpu-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl\nPython 3.12 GPU support\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_x86_64.whl\nPython 3.12 CPU-only\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/2.20.0\/tensorflow_cpu-2.20.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl\nPython 3.13 GPU support\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl\nPython 3.13 CPU-only\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow_cpu-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl\nLinux Arm64 (CPU-only)\nPython 3.10\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_aarch64.whl\nPython 3.11\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_aarch64.whl\nPython 3.12\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_aarch64.whl\nPython 3.13\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_aarch64.whl\nmacOS x86 (CPU-only)\nCaution\n: TensorFlow 2.16 was the\nlast\nTensorFlow release that supported macOS x86\nPython 3.10\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/2.16.2\/tensorflow-2.16.2-cp310-cp310-macosx_10_15_x86_64.whl\nPython 3.11\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/2.16.2\/tensorflow-2.16.2-cp311-cp311-macosx_10_15_x86_64.whl\nPython 3.12\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/2.16.2\/tensorflow-2.16.2-cp312-cp312-macosx_10_15_x86_64.whl\nmacOS Arm64 (CPU-only)\nPython 3.10\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-macosx_12_0_arm64.whl\nPython 3.11\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-macosx_12_0_arm64.whl\nPython 3.12\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-macosx_12_0_arm64.whl\nPython 3.13\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-macosx_12_0_arm64.whl\nWindows (CPU-only)\nPython 3.10\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-win_amd64.whl\nPython 3.11\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-win_amd64.whl\nPython 3.12\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-win_amd64.whl\nPython 3.13\nhttps:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-win_amd64.whl\nExcept as otherwise noted, the content of this page is licensed under the\nCreative Commons Attribution 4.0 License\n, and code samples are licensed under the\nApache 2.0 License\n. For details, see the\nGoogle Developers Site Policies\n. Java is a registered trademark of Oracle and\/or its affiliates.\nLast updated 2026-03-12 UTC.\n[[[\"Easy to understand\",\"easyToUnderstand\",\"thumb-up\"],[\"Solved my problem\",\"solvedMyProblem\",\"thumb-up\"],[\"Other\",\"otherUp\",\"thumb-up\"]],[[\"Missing the information I need\",\"missingTheInformationINeed\",\"thumb-down\"],[\"Too complicated \/ too many steps\",\"tooComplicatedTooManySteps\",\"thumb-down\"],[\"Out of date\",\"outOfDate\",\"thumb-down\"],[\"Samples \/ code issue\",\"samplesCodeIssue\",\"thumb-down\"],[\"Other\",\"otherDown\",\"thumb-down\"]],[\"Last updated 2026-03-12 UTC.\"],[],[]]","attrs_markdown":"[Skip to main content](https:\/\/www.tensorflow.org\/install\/pip#main-content)\n\n[![TensorFlow](https:\/\/www.gstatic.com\/devrel-devsite\/prod\/v369eac9380f92e8fedc492e2689927bb3475d758266c381eee326d0b49a77481\/tensorflow\/images\/lockup.svg)](https:\/\/www.tensorflow.org\/)\n\n[Install](https:\/\/www.tensorflow.org\/install)\n\n[Learn](https:\/\/www.tensorflow.org\/learn)\n\n- [Introduction New to TensorFlow?](https:\/\/www.tensorflow.org\/learn)\n- [Tutorials Learn how to use TensorFlow with end-to-end examples](https:\/\/www.tensorflow.org\/tutorials)\n- [Guide Learn framework concepts and components](https:\/\/www.tensorflow.org\/guide)\n- [Learn ML Educational resources to master your path with TensorFlow](https:\/\/www.tensorflow.org\/resources\/learn-ml)\n\nMore\n\n[API](https:\/\/www.tensorflow.org\/api)\n\n- [TensorFlow (v2.16.1)](https:\/\/www.tensorflow.org\/api\/stable)\n- [Versions…](https:\/\/www.tensorflow.org\/versions)\n\n- [TensorFlow.js](https:\/\/js.tensorflow.org\/api\/latest\/)\n- [TensorFlow Lite](https:\/\/www.tensorflow.org\/lite\/api_docs)\n- [TFX](https:\/\/www.tensorflow.org\/tfx\/api_docs)\n\n[Ecosystem](https:\/\/www.tensorflow.org\/resources\/models-datasets)\n\n- LIBRARIES\n- [TensorFlow.js Develop web ML applications in JavaScript](https:\/\/www.tensorflow.org\/js)\n- [TensorFlow Lite Deploy ML on mobile, microcontrollers and other edge devices](https:\/\/www.tensorflow.org\/lite)\n- [TFX Build production ML pipelines](https:\/\/www.tensorflow.org\/tfx)\n- [All libraries Create advanced models and extend TensorFlow](https:\/\/www.tensorflow.org\/resources\/libraries-extensions)\n\n- RESOURCES\n- [Models & datasets Pre-trained models and datasets built by Google and the community](https:\/\/www.tensorflow.org\/resources\/models-datasets)\n- [Tools Tools to support and accelerate TensorFlow workflows](https:\/\/www.tensorflow.org\/resources\/tools)\n- [Responsible AI Resources for every stage of the ML workflow](https:\/\/www.tensorflow.org\/responsible_ai)\n- [Recommendation systems Build recommendation systems with open source tools](https:\/\/www.tensorflow.org\/resources\/recommendation-systems)\n\n[Community](https:\/\/www.tensorflow.org\/community)\n\n- [Groups User groups, interest groups and mailing lists](https:\/\/www.tensorflow.org\/community\/groups)\n- [Contribute Guide for contributing to code and documentation](https:\/\/www.tensorflow.org\/community\/contribute)\n- [Blog Stay up to date with all things TensorFlow](https:\/\/blog.tensorflow.org\/)\n- [Forum Discussion platform for the TensorFlow community](https:\/\/discuss.tensorflow.org\/)\n\n[Why TensorFlow](https:\/\/www.tensorflow.org\/about)\n\n- [About](https:\/\/www.tensorflow.org\/about)\n- [Case studies](https:\/\/www.tensorflow.org\/about\/case-studies)\n\n- [English](https:\/\/www.tensorflow.org\/install\/pip)\n- [Español](https:\/\/www.tensorflow.org\/install\/pip?hl=es)\n- [Español – América Latina](https:\/\/www.tensorflow.org\/install\/pip?hl=es-419)\n- [Français](https:\/\/www.tensorflow.org\/install\/pip?hl=fr)\n- [Indonesia](https:\/\/www.tensorflow.org\/install\/pip?hl=id)\n- [Italiano](https:\/\/www.tensorflow.org\/install\/pip?hl=it)\n- [Polski](https:\/\/www.tensorflow.org\/install\/pip?hl=pl)\n- [Português](https:\/\/www.tensorflow.org\/install\/pip?hl=pt)\n- [Português – Brasil](https:\/\/www.tensorflow.org\/install\/pip?hl=pt-br)\n- [Tiếng Việt](https:\/\/www.tensorflow.org\/install\/pip?hl=vi)\n- [Türkçe](https:\/\/www.tensorflow.org\/install\/pip?hl=tr)\n- [Русский](https:\/\/www.tensorflow.org\/install\/pip?hl=ru)\n- [עברית](https:\/\/www.tensorflow.org\/install\/pip?hl=he)\n- [العربيّة](https:\/\/www.tensorflow.org\/install\/pip?hl=ar)\n- [فارسی](https:\/\/www.tensorflow.org\/install\/pip?hl=fa)\n- [हिंदी](https:\/\/www.tensorflow.org\/install\/pip?hl=hi)\n- [বাংলা](https:\/\/www.tensorflow.org\/install\/pip?hl=bn)\n- [ภาษาไทย](https:\/\/www.tensorflow.org\/install\/pip?hl=th)\n- [中文 – 简体](https:\/\/www.tensorflow.org\/install\/pip?hl=zh-cn)\n- [中文 – 繁體](https:\/\/www.tensorflow.org\/install\/pip?hl=zh-tw)\n- [日本語](https:\/\/www.tensorflow.org\/install\/pip?hl=ja)\n- [한국어](https:\/\/www.tensorflow.org\/install\/pip?hl=ko)\n\n[GitHub](https:\/\/github.com\/tensorflow)\n\n[Sign in](https:\/\/www.tensorflow.org\/_d\/signin?continue=https%3A%2F%2Fwww.tensorflow.org%2Finstall%2Fpip&prompt=select_account)\n\n- [Install](https:\/\/www.tensorflow.org\/install)\n\n[![TensorFlow](https:\/\/www.gstatic.com\/devrel-devsite\/prod\/v369eac9380f92e8fedc492e2689927bb3475d758266c381eee326d0b49a77481\/tensorflow\/images\/lockup.svg)](https:\/\/www.tensorflow.org\/)\n\n- [Install](https:\/\/www.tensorflow.org\/install)\n- [Learn](https:\/\/www.tensorflow.org\/learn)\n  - More\n- [API](https:\/\/www.tensorflow.org\/api)\n  - More\n- [Ecosystem](https:\/\/www.tensorflow.org\/resources\/models-datasets)\n  - More\n- [Community](https:\/\/www.tensorflow.org\/community)\n  - More\n- [Why TensorFlow](https:\/\/www.tensorflow.org\/about)\n  - More\n- [GitHub](https:\/\/github.com\/tensorflow)\n\n- [Install Tensor Flow](https:\/\/www.tensorflow.org\/install)\n- Packages\n- [pip](https:\/\/www.tensorflow.org\/install\/pip)\n- [Docker](https:\/\/www.tensorflow.org\/install\/docker)\n- Additional setup\n- [GPU device plugins](https:\/\/www.tensorflow.org\/install\/gpu_plugins)\n- [Problems](https:\/\/www.tensorflow.org\/install\/errors)\n- Build from source\n- [Linux \/ mac OS](https:\/\/www.tensorflow.org\/install\/source)\n- [Windows](https:\/\/www.tensorflow.org\/install\/source_windows)\n- [SIG Build](https:\/\/github.com\/tensorflow\/build)\n- Language bindings\n- [Java](https:\/\/www.tensorflow.org\/jvm\/install)\n- [Java (legacy)](https:\/\/www.tensorflow.org\/install\/lang_java_legacy)\n- [C](https:\/\/www.tensorflow.org\/install\/lang_c)\n- [Go](https:\/\/github.com\/tensorflow\/build\/tree\/master\/golang_install_guide)\n\n- [Introduction](https:\/\/www.tensorflow.org\/learn)\n- [Tutorials](https:\/\/www.tensorflow.org\/tutorials)\n- [Guide](https:\/\/www.tensorflow.org\/guide)\n- [Learn ML](https:\/\/www.tensorflow.org\/resources\/learn-ml)\n\n- [TensorFlow (v2.16.1)](https:\/\/www.tensorflow.org\/api\/stable)\n- [Versions…](https:\/\/www.tensorflow.org\/versions)\n- [TensorFlow.js](https:\/\/js.tensorflow.org\/api\/latest\/)\n- [TensorFlow Lite](https:\/\/www.tensorflow.org\/lite\/api_docs)\n- [TFX](https:\/\/www.tensorflow.org\/tfx\/api_docs)\n\n- LIBRARIES\n- [TensorFlow.js](https:\/\/www.tensorflow.org\/js)\n- [TensorFlow Lite](https:\/\/www.tensorflow.org\/lite)\n- [TFX](https:\/\/www.tensorflow.org\/tfx)\n- [All libraries](https:\/\/www.tensorflow.org\/resources\/libraries-extensions)\n- RESOURCES\n- [Models & datasets](https:\/\/www.tensorflow.org\/resources\/models-datasets)\n- [Tools](https:\/\/www.tensorflow.org\/resources\/tools)\n- [Responsible AI](https:\/\/www.tensorflow.org\/responsible_ai)\n- [Recommendation systems](https:\/\/www.tensorflow.org\/resources\/recommendation-systems)\n\n- [Groups](https:\/\/www.tensorflow.org\/community\/groups)\n- [Contribute](https:\/\/www.tensorflow.org\/community\/contribute)\n- [Blog](https:\/\/blog.tensorflow.org\/)\n- [Forum](https:\/\/discuss.tensorflow.org\/)\n\n- [About](https:\/\/www.tensorflow.org\/about)\n- [Case studies](https:\/\/www.tensorflow.org\/about\/case-studies)\n\n- On this page\n- [Hardware requirements](https:\/\/www.tensorflow.org\/install\/pip#hardware_requirements)\n- [System requirements](https:\/\/www.tensorflow.org\/install\/pip#system_requirements)\n- [Software requirements](https:\/\/www.tensorflow.org\/install\/pip#software_requirements)\n- [Step-by-step instructions](https:\/\/www.tensorflow.org\/install\/pip#step-by-step_instructions)\n- [Package location](https:\/\/www.tensorflow.org\/install\/pip#package_location)\n- [Python version support](https:\/\/www.tensorflow.org\/install\/pip#python_version_support)\n\n- [TensorFlow](https:\/\/www.tensorflow.org\/)\n- [Install](https:\/\/www.tensorflow.org\/install)\n\nWas this helpful?\n# Install TensorFlow with pipStay organized with collections Save and categorize content based on your preferences.\n- On this page\n- [Hardware requirements](https:\/\/www.tensorflow.org\/install\/pip#hardware_requirements)\n- [System requirements](https:\/\/www.tensorflow.org\/install\/pip#system_requirements)\n- [Software requirements](https:\/\/www.tensorflow.org\/install\/pip#software_requirements)\n- [Step-by-step instructions](https:\/\/www.tensorflow.org\/install\/pip#step-by-step_instructions)\n- [Package location](https:\/\/www.tensorflow.org\/install\/pip#package_location)\n- [Python version support](https:\/\/www.tensorflow.org\/install\/pip#python_version_support)\n\nThis guide is for the latest stable version of TensorFlow. For the preview build *(nightly)*, use the pip package named `tf-nightly`. Refer to [these tables](https:\/\/www.tensorflow.org\/install\/source#tested_build_configurations) for older TensorFlow version requirements. For the CPU-only build, use the pip package named `tensorflow-cpu`.\n\nHere are the quick versions of the install commands. Scroll down for the step-by-step instructions.\n\n[Linux](https:\/\/www.tensorflow.org\/install\/pip#linux)\n\n[MacOS](https:\/\/www.tensorflow.org\/install\/pip#macos)\n\n[Windows Native](https:\/\/www.tensorflow.org\/install\/pip#windows-native)\n\n[Windows WSL2](https:\/\/www.tensorflow.org\/install\/pip#windows-wsl2)\n\n[CPU](https:\/\/www.tensorflow.org\/install\/pip#cpu)\n\n[Nightly](https:\/\/www.tensorflow.org\/install\/pip#nightly)\n\nMore\n\n**Note:** Starting with TensorFlow `2.10`, Linux CPU-builds for Aarch64\/ARM64 processors are built, maintained, tested and released by a third party: [AWS](https:\/\/aws.amazon.com\/). Installing the [`tensorflow`](https:\/\/pypi.org\/project\/tensorflow\/) package on an ARM machine installs AWS's [`tensorflow-cpu-aws`](https:\/\/pypi.org\/project\/tensorflow-cpu-aws\/) package. They are provided as-is. Tensorflow will use reasonable efforts to maintain the availability and integrity of this pip package. There may be delays if the third party fails to release the pip package. See [this blog post](https:\/\/blog.tensorflow.org\/2022\/09\/announcing-tensorflow-official-build-collaborators.html) for more information about this collaboration.\n```\npython3 -m pip install 'tensorflow[and-cuda]'\n# Verify the installation:\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\n```\n# There is currently no official GPU support for MacOS.\npython3 -m pip install tensorflow\n# Verify the installation:\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\n**Caution:** TensorFlow `2.10` was the **last** TensorFlow release that supported GPU on native-Windows. Starting with TensorFlow `2.11`, you will need to install [TensorFlow in WSL2](https:\/\/tensorflow.org\/install\/pip#windows-wsl2), or install `tensorflow` or `tensorflow-cpu` and, optionally, try the [TensorFlow-DirectML-Plugin](https:\/\/github.com\/microsoft\/tensorflow-directml-plugin#tensorflow-directml-plugin-)\n```\nconda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0\n# Anything above 2.10 is not supported on the GPU on Windows Native\npython -m pip install \"tensorflow<2.11\"\n# Verify the installation:\npython -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\n\n**Note:** TensorFlow with GPU access is supported for WSL2 on Windows 10 19044 or higher. This corresponds to Windows 10 version 21H2, the November 2021 update. You can get the latest update from here: [Download Windows 10](https:\/\/www.microsoft.com\/software-download\/windows10). For instructions, see [Install WSL2](https:\/\/docs.microsoft.com\/windows\/wsl\/install) and [NVIDIA’s setup docs](https:\/\/docs.nvidia.com\/cuda\/wsl-user-guide\/index.html) for CUDA in WSL.\n```\npython3 -m pip install tensorflow[and-cuda]\n# Verify the installation:\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\n\n**Note:** Starting with TensorFlow `2.10`, Windows CPU-builds for x86\/x64 processors are built, maintained, tested and released by a third party: [Intel](https:\/\/www.intel.com\/). Installing the Windows-native [`tensorflow`](https:\/\/pypi.org\/project\/tensorflow\/) or [`tensorflow-cpu`](https:\/\/pypi.org\/project\/tensorflow-cpu\/) package installs Intel's [`tensorflow-intel`](https:\/\/pypi.org\/project\/tensorflow-intel\/) package. These packages are provided as-is. Tensorflow will use reasonable efforts to maintain the availability and integrity of this pip package. There may be delays if the third party fails to release the pip package. See [this blog post](https:\/\/blog.tensorflow.org\/2022\/09\/announcing-tensorflow-official-build-collaborators.html) for more information about this collaboration.\n```\npython3 -m pip install tensorflow\n# Verify the installation:\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\n```\npython3 -m pip install tf-nightly\n# Verify the installation:\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\n## Hardware requirements\n**Note:** TensorFlow binaries use [AVX instructions](https:\/\/en.wikipedia.org\/wiki\/Advanced_Vector_Extensions#CPUs_with_AVX) which may not run on older CPUs.\n\nThe following GPU-enabled devices are supported:\n\n- NVIDIA® GPU card with CUDA® architectures 3.5, 5.0, 6.0, 7.0, 7.5, 8.0 and higher. See the list of [CUDA®-enabled GPU cards](https:\/\/developer.nvidia.com\/cuda-gpus).\n- For GPUs with unsupported CUDA® architectures, or to avoid JIT compilation from PTX, or to use different versions of the NVIDIA® libraries, see the [Linux build from source](https:\/\/www.tensorflow.org\/install\/source) guide.\n- Packages do not contain PTX code except for the latest supported CUDA® architecture; therefore, TensorFlow fails to load on older GPUs when `CUDA_FORCE_PTX_JIT=1` is set. (See [Application Compatibility](https:\/\/docs.nvidia.com\/cuda\/cuda-c-programming-guide\/index.html#application-compatibility) for details.)\n\n**Note:** The error message \"Status: device kernel image is invalid\" indicates that the TensorFlow package does not contain PTX for your architecture. You can enable compute capabilities by [building TensorFlow from source](https:\/\/www.tensorflow.org\/install\/source).\n\n## System requirements\n- Ubuntu 16.04 or higher (64-bit)\n- macOS 12.0 (Monterey) or higher (64-bit) *(no GPU support)*\n- Windows Native - Windows 7 or higher (64-bit) *(no GPU support after TF 2.10)*\n- Windows WSL2 - Windows 10 19044 or higher (64-bit)\n\n**Note:** GPU support is available for Ubuntu and Windows with CUDA®-enabled cards.\n\n## Software requirements\n- Python 3.9–3.12\n- pip version 19.0 or higher for Linux (requires `manylinux2014` support) and Windows. pip version 20.3 or higher for macOS.\n- Windows Native Requires [Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019](https:\/\/learn.microsoft.com\/en-us\/cpp\/windows\/latest-supported-vc-redist)\n\nThe following NVIDIA® software are only required for GPU support.\n\n- [NVIDIA® GPU drivers](https:\/\/www.nvidia.com\/drivers)\n  - \\>= 525.60.13 for Linux\n  - \\>= 528.33 for WSL on Windows\n- [CUDA® Toolkit 12.3](https:\/\/developer.nvidia.com\/cuda-toolkit-archive).\n- [cuDNN SDK 8.9.7](https:\/\/developer.nvidia.com\/cudnn).\n- *(Optional)* [TensorRT](https:\/\/docs.nvidia.com\/deeplearning\/tensorrt\/archives\/index.html#trt_7) to improve latency and throughput for inference.\n## Step-by-step instructions\n[Linux](https:\/\/www.tensorflow.org\/install\/pip#linux)\n\n[MacOS](https:\/\/www.tensorflow.org\/install\/pip#macos)\n\n[Windows Native](https:\/\/www.tensorflow.org\/install\/pip#windows-native)\n\n[Windows WSL2](https:\/\/www.tensorflow.org\/install\/pip#windows-wsl2)\n\nMore\n\n### 1\\. System requirements\n- Ubuntu 16.04 or higher (64-bit)\n\nTensorFlow only officially supports Ubuntu. However, the following instructions may also work for other Linux distros.\n\n**Note:** Starting with TensorFlow `2.10`, Linux CPU-builds for Aarch64\/ARM64 processors are built, maintained, tested and released by a third party: [AWS](https:\/\/aws.amazon.com\/). Installing the [`tensorflow`](https:\/\/pypi.org\/project\/tensorflow\/) package on an ARM machine installs AWS's [`tensorflow-cpu-aws`](https:\/\/pypi.org\/project\/tensorflow-cpu-aws\/) package. They are provided as-is. Tensorflow will use reasonable efforts to maintain the availability and integrity of this pip package. There may be delays if the third party fails to release the pip package. See [this blog post](https:\/\/blog.tensorflow.org\/2022\/09\/announcing-tensorflow-official-build-collaborators.html) for more information about this collaboration.\n\n### 2\\. GPU setup\nYou can skip this section if you only run TensorFlow on the CPU.\n\nInstall the [NVIDIA GPU driver](https:\/\/www.nvidia.com\/Download\/index.aspx) if you have not. You can use the following command to verify it is installed.\n```\nnvidia-smi\n```\n### 3\\. Create a virtual environment with [venv](https:\/\/docs.python.org\/3\/library\/venv.html)\nThe venv module is part of Python’s standard library and is the officially recommended way to create virtual environments.\n\nNavigate to your desired virtual environments directory and create a new venv environment named [`tf`](https:\/\/www.tensorflow.org\/api_docs\/python\/tf) with the following command.\n```\npython3 -m venv tf\n```\nYou can activate it with the following command.\n```\nsource tf\/bin\/activate\n```\nMake sure that the virtual environment is activated for the rest of the installation.\n\n### 4\\. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.\n```\npip install --upgrade pip\n```\nThen, install TensorFlow with pip.\n```\n# For GPU users\npip install tensorflow[and-cuda]\n# For CPU users\npip install tensorflow\n```\n**Note:** Do not install TensorFlow with `conda`. It may not have the latest stable version. `pip` is recommended since TensorFlow is only officially released to PyPI.\n\n### 6\\. Verify the installation\nVerify the CPU setup:\n```\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\nIf a tensor is returned, you've installed TensorFlow successfully.\n\nVerify the GPU setup:\n```\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\nIf a list of GPU devices is returned, you've installed TensorFlow successfully. **If not continue to the next step**.\n\n### 6\\. \\[GPU only\\] Virtual environment configuration\nIf the GPU test in the last section was unsuccessful, the most likely cause is that components aren't being detected, and\/or conflict with the existing system CUDA installation. So you need to add some symbolic links to fix this.\n\n- Create symbolic links to NVIDIA shared libraries:\n```\npushd $(dirname $(python -c 'print(__import__(\"tensorflow\").__file__)'))\nln -svf ..\/nvidia\/*\/lib\/*.so* .\npopd\n```\n- Create a symbolic link to ptxas:\n```\nln -sf $(find $(dirname $(dirname $(python -c \"import nvidia.cuda_nvcc;         \nprint(nvidia.cuda_nvcc.__file__)\"))\/*\/bin\/) -name ptxas -print -quit) $VIRTUAL_ENV\/bin\/ptxas\n```\nVerify the GPU setup:\n```\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\n\n### 1\\. System requirements\n- macOS 10.12.6 (Sierra) or higher (64-bit)\n\n**Note:** While TensorFlow supports Apple Silicon (M1), packages that include custom C++ extensions for TensorFlow also need to be compiled for Apple M1. Some packages, like [tensorflow\\_decision\\_forests](https:\/\/www.tensorflow.org\/decision_forests) publish M1-compatible versions, but many packages don't. To use those libraries, you will have to use TensorFlow with x86 emulation and Rosetta.\n\nCurrently there is no official GPU support for running TensorFlow on MacOS. The following instructions are for running on CPU.\n\n### 2\\. Check Python version\nCheck if your Python environment is already configured:\n\n**Note:** Requires Python 3.9–3.11, and pip \\>= 20.3 for MacOS.\n```\npython3 --version\npython3 -m pip --version\n```\n### 3\\. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.\n```\npip install --upgrade pip\n```\nThen, install TensorFlow with pip.\n```\npip install tensorflow\n```\n### 4\\. Verify the installation\n```\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\nIf a tensor is returned, you've installed TensorFlow successfully.\n\n**Caution:** TensorFlow `2.10` was the **last** TensorFlow release that supported GPU on native-Windows. Starting with TensorFlow `2.11`, you will need to install [TensorFlow in WSL2](https:\/\/tensorflow.org\/install\/pip#windows-%5Bwsl2%5D), or install `tensorflow-cpu` and, optionally, try the [TensorFlow-DirectML-Plugin](https:\/\/github.com\/microsoft\/tensorflow-directml-plugin#tensorflow-directml-plugin-)\n\n## 1\\. System requirements\n- Windows 7 or higher (64-bit)\n\n**Note:** Starting with TensorFlow `2.10`, Windows CPU-builds for x86\/x64 processors are built, maintained, tested and released by a third party: [Intel](https:\/\/www.intel.com\/). Installing the windows-native [`tensorflow`](https:\/\/pypi.org\/project\/tensorflow\/) or [`tensorflow-cpu`](https:\/\/pypi.org\/project\/tensorflow-cpu\/) package installs Intel's [`tensorflow-intel`](https:\/\/pypi.org\/project\/tensorflow-intel\/) package. These packages are provided as-is. Tensorflow will use reasonable efforts to maintain the availability and integrity of this pip package. There may be delays if the third party fails to release the pip package. See [this blog post](https:\/\/blog.tensorflow.org\/2022\/09\/announcing-tensorflow-official-build-collaborators.html) for more information about this collaboration.\n\n### 2\\. Install Microsoft Visual C++ Redistributable\nInstall the *Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017, and 2019*. Starting with the TensorFlow 2.1.0 version, the `msvcp140_1.dll` file is required from this package (which may not be provided from older redistributable packages). The redistributable comes with *Visual Studio 2019* but can be installed separately:\n\n1. Go to the [Microsoft Visual C++ downloads](https:\/\/support.microsoft.com\/help\/2977003\/the-latest-supported-visual-c-downloads).\n2. Scroll down the page to the *Visual Studio 2015, 2017 and 2019* section.\n3. Download and install the *Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019* for your platform.\n\nMake sure [long paths are enabled](https:\/\/superuser.com\/questions\/1119883\/windows-10-enable-ntfs-long-paths-policy-option-missing) on Windows.\n\n### 3\\. Install Miniconda\n[Miniconda](https:\/\/docs.conda.io\/en\/latest\/miniconda.html) is the recommended approach for installing TensorFlow with GPU support. It creates a separate environment to avoid changing any installed software in your system. This is also the easiest way to install the required software especially for the GPU setup.\n\nDownload the [Miniconda Windows Installer](https:\/\/repo.anaconda.com\/miniconda\/Miniconda3-latest-Windows-x86_64.exe). Double-click the downloaded file and follow the instructions on the screen.\n\n### 4\\. Create a conda environment\nCreate a new conda environment named [`tf`](https:\/\/www.tensorflow.org\/api_docs\/python\/tf) with the following command.\n```\nconda create --name tf python=3.9\n```\nYou can deactivate and activate it with the following commands.\n```\nconda deactivate\nconda activate tf\n```\nMake sure it is activated for the rest of the installation.\n\n### 5\\. GPU setup\nYou can skip this section if you only run TensorFlow on CPU.\n\nFirst install [NVIDIA GPU driver](https:\/\/www.nvidia.com\/Download\/index.aspx) if you have not.\n\nThen install the CUDA, cuDNN with conda.\n```\nconda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0\n```\n### 6\\. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.\n```\npip install --upgrade pip\n```\nThen, install TensorFlow with pip.\n\n**Note:** Do not install TensorFlow with conda. It may not have the latest stable version. pip is recommended since TensorFlow is only officially released to PyPI.\n```\n# Anything above 2.10 is not supported on the GPU on Windows Native\npip install \"tensorflow<2.11\"\n```\n### 7\\. Verify the installation\nVerify the CPU setup:\n```\npython -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\nIf a tensor is returned, you've installed TensorFlow successfully.\n\nVerify the GPU setup:\n```\npython -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\nIf a list of GPU devices is returned, you've installed TensorFlow successfully.\n\n### 1\\. System requirements\n- Windows 10 19044 or higher (64-bit). This corresponds to Windows 10 version 21H2, the November 2021 update.\n\nSee the following documents to:\n\n- [Download the latest Windows 10 update](https:\/\/www.microsoft.com\/software-download\/windows10).\n- [Install WSL2](https:\/\/docs.microsoft.com\/windows\/wsl\/install)\n- [Setup NVIDIA® GPU support in WSL2](https:\/\/docs.nvidia.com\/cuda\/wsl-user-guide\/index.html)\n### 2\\. GPU setup\nYou can skip this section if you only run TensorFlow on the CPU.\n\nInstall the [NVIDIA GPU driver](https:\/\/www.nvidia.com\/Download\/index.aspx) if you have not. You can use the following command to verify it is installed.\n```\nnvidia-smi\n```\n### 3\\. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.\n```\npip install --upgrade pip\n```\nThen, install TensorFlow with pip.\n```\n# For GPU users\npip install tensorflow[and-cuda]\n# For CPU users\npip install tensorflow\n```\n### 4\\. Verify the installation\nVerify the CPU setup:\n```\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\nIf a tensor is returned, you've installed TensorFlow successfully.\n\nVerify the GPU setup:\n```\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\nIf a list of GPU devices is returned, you've installed TensorFlow successfully.\n## Package location\nA few installation mechanisms require the URL of the TensorFlow Python package. The value you specify depends on your Python version.\n\n## Python version support\n**Warning:** As of TensorFlow 2.21, Python 3.9 is no longer supported. Please use a supported Python version (e.g., 3.10-3.13).\n\n| Version | URL |\n|---|---|\n| Linux x86 |   |\n| Python 3.10 GPU support | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl) |\n| Python 3.10 CPU-only | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow\\_cpu-2.21.0-cp310-cp310-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow_cpu-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl) |\n| Python 3.11 GPU support | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl) |\n| Python 3.11 CPU-only | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow\\_cpu-2.21.0-cp311-cp311-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow_cpu-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl) |\n| Python 3.12 GPU support | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_x86_64.whl) |\n| Python 3.12 CPU-only | <https:\/\/storage.googleapis.com\/tensorflow\/versions\/2.20.0\/tensorflow_cpu-2.20.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl> |\n| Python 3.13 GPU support | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl) |\n| Python 3.13 CPU-only | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow\\_cpu-2.21.0-cp313-cp313-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow_cpu-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl) |\n| Linux Arm64 (CPU-only) |   |\n| Python 3.10 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-manylinux\\_2\\_27\\_aarch64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_aarch64.whl) |\n| Python 3.11 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-manylinux\\_2\\_27\\_aarch64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_aarch64.whl) |\n| Python 3.12 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-manylinux\\_2\\_27\\_aarch64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_aarch64.whl) |\n| Python 3.13 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-manylinux\\_2\\_27\\_aarch64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_aarch64.whl) |\n| macOS x86 (CPU-only) |   |\n| **Caution**: TensorFlow 2.16 was the **last** TensorFlow release that supported macOS x86 |   |\n| Python 3.10 | <https:\/\/storage.googleapis.com\/tensorflow\/versions\/2.16.2\/tensorflow-2.16.2-cp310-cp310-macosx_10_15_x86_64.whl> |\n| Python 3.11 | <https:\/\/storage.googleapis.com\/tensorflow\/versions\/2.16.2\/tensorflow-2.16.2-cp311-cp311-macosx_10_15_x86_64.whl> |\n| Python 3.12 | <https:\/\/storage.googleapis.com\/tensorflow\/versions\/2.16.2\/tensorflow-2.16.2-cp312-cp312-macosx_10_15_x86_64.whl> |\n| macOS Arm64 (CPU-only) |   |\n| Python 3.10 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-macosx\\_12\\_0\\_arm64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-macosx_12_0_arm64.whl) |\n| Python 3.11 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-macosx\\_12\\_0\\_arm64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-macosx_12_0_arm64.whl) |\n| Python 3.12 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-macosx\\_12\\_0\\_arm64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-macosx_12_0_arm64.whl) |\n| Python 3.13 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-macosx\\_12\\_0\\_arm64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-macosx_12_0_arm64.whl) |\n| Windows (CPU-only) |   |\n| Python 3.10 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-win\\_amd64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-win_amd64.whl) |\n| Python 3.11 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-win\\_amd64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-win_amd64.whl) |\n| Python 3.12 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-win\\_amd64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-win_amd64.whl) |\n| Python 3.13 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-win\\_amd64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-win_amd64.whl) |\n\nWas this helpful?\n\nExcept as otherwise noted, the content of this page is licensed under the [Creative Commons Attribution 4.0 License](https:\/\/creativecommons.org\/licenses\/by\/4.0\/), and code samples are licensed under the [Apache 2.0 License](https:\/\/www.apache.org\/licenses\/LICENSE-2.0). For details, see the [Google Developers Site Policies](https:\/\/developers.google.com\/site-policies). Java is a registered trademark of Oracle and\/or its affiliates.\n\nLast updated 2026-03-12 UTC.\n\n- ### Stay connected\n  - [Blog](https:\/\/blog.tensorflow.org\/)\n  - [Forum](https:\/\/discuss.tensorflow.org\/)\n  - [GitHub](https:\/\/github.com\/tensorflow\/)\n  - [Twitter](https:\/\/twitter.com\/tensorflow)\n  - [YouTube](https:\/\/youtube.com\/tensorflow)\n- ### Support\n  - [Issue tracker](https:\/\/github.com\/tensorflow\/tensorflow\/issues)\n  - [Release notes](https:\/\/github.com\/tensorflow\/tensorflow\/blob\/master\/RELEASE.md)\n  - [Stack Overflow](https:\/\/stackoverflow.com\/questions\/tagged\/tensorflow)\n  - [Brand guidelines](https:\/\/www.tensorflow.org\/extras\/tensorflow_brand_guidelines.pdf)\n  - [Cite TensorFlow](https:\/\/www.tensorflow.org\/about\/bib)\n\n- [Terms](https:\/\/policies.google.com\/terms)\n- [Privacy](https:\/\/policies.google.com\/privacy)\n- [Manage cookies](https:\/\/www.tensorflow.org\/install\/pip)\n- Sign up for the TensorFlow newsletter  [Subscribe](https:\/\/www.tensorflow.org\/subscribe)\n\n- [English](https:\/\/www.tensorflow.org\/install\/pip)\n- [Español](https:\/\/www.tensorflow.org\/install\/pip?hl=es)\n- [Español – América Latina](https:\/\/www.tensorflow.org\/install\/pip?hl=es-419)\n- [Français](https:\/\/www.tensorflow.org\/install\/pip?hl=fr)\n- [Indonesia](https:\/\/www.tensorflow.org\/install\/pip?hl=id)\n- [Italiano](https:\/\/www.tensorflow.org\/install\/pip?hl=it)\n- [Polski](https:\/\/www.tensorflow.org\/install\/pip?hl=pl)\n- [Português](https:\/\/www.tensorflow.org\/install\/pip?hl=pt)\n- [Português – Brasil](https:\/\/www.tensorflow.org\/install\/pip?hl=pt-br)\n- [Tiếng Việt](https:\/\/www.tensorflow.org\/install\/pip?hl=vi)\n- [Türkçe](https:\/\/www.tensorflow.org\/install\/pip?hl=tr)\n- [Русский](https:\/\/www.tensorflow.org\/install\/pip?hl=ru)\n- [עברית](https:\/\/www.tensorflow.org\/install\/pip?hl=he)\n- [العربيّة](https:\/\/www.tensorflow.org\/install\/pip?hl=ar)\n- [فارسی](https:\/\/www.tensorflow.org\/install\/pip?hl=fa)\n- [हिंदी](https:\/\/www.tensorflow.org\/install\/pip?hl=hi)\n- [বাংলা](https:\/\/www.tensorflow.org\/install\/pip?hl=bn)\n- [ภาษาไทย](https:\/\/www.tensorflow.org\/install\/pip?hl=th)\n- [中文 – 简体](https:\/\/www.tensorflow.org\/install\/pip?hl=zh-cn)\n- [中文 – 繁體](https:\/\/www.tensorflow.org\/install\/pip?hl=zh-tw)\n- [日本語](https:\/\/www.tensorflow.org\/install\/pip?hl=ja)\n- [한국어](https:\/\/www.tensorflow.org\/install\/pip?hl=ko)","attrs_readable_markdown":"## Install TensorFlow with pipStay organized with collections Save and categorize content based on your preferences.\n- On this page\n- [Hardware requirements](https:\/\/www.tensorflow.org\/install\/pip#hardware_requirements)\n- [System requirements](https:\/\/www.tensorflow.org\/install\/pip#system_requirements)\n- [Software requirements](https:\/\/www.tensorflow.org\/install\/pip#software_requirements)\n- [Step-by-step instructions](https:\/\/www.tensorflow.org\/install\/pip#step-by-step_instructions)\n- [Package location](https:\/\/www.tensorflow.org\/install\/pip#package_location)\n- [Python version support](https:\/\/www.tensorflow.org\/install\/pip#python_version_support)\n\nThis guide is for the latest stable version of TensorFlow. For the preview build *(nightly)*, use the pip package named `tf-nightly`. Refer to [these tables](https:\/\/www.tensorflow.org\/install\/source#tested_build_configurations) for older TensorFlow version requirements. For the CPU-only build, use the pip package named `tensorflow-cpu`.\n\nHere are the quick versions of the install commands. Scroll down for the step-by-step instructions.\n\n```\npython3 -m pip install 'tensorflow[and-cuda]'\n# Verify the installation:\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\n```\n# There is currently no official GPU support for MacOS.\npython3 -m pip install tensorflow\n# Verify the installation:\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\n```\nconda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0\n# Anything above 2.10 is not supported on the GPU on Windows Native\npython -m pip install \"tensorflow<2.11\"\n# Verify the installation:\npython -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\n```\npython3 -m pip install tensorflow[and-cuda]\n# Verify the installation:\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\n```\npython3 -m pip install tensorflow\n# Verify the installation:\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\n```\npython3 -m pip install tf-nightly\n# Verify the installation:\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\n## Hardware requirements\nThe following GPU-enabled devices are supported:\n\n- NVIDIA® GPU card with CUDA® architectures 3.5, 5.0, 6.0, 7.0, 7.5, 8.0 and higher. See the list of [CUDA®-enabled GPU cards](https:\/\/developer.nvidia.com\/cuda-gpus).\n- For GPUs with unsupported CUDA® architectures, or to avoid JIT compilation from PTX, or to use different versions of the NVIDIA® libraries, see the [Linux build from source](https:\/\/www.tensorflow.org\/install\/source) guide.\n- Packages do not contain PTX code except for the latest supported CUDA® architecture; therefore, TensorFlow fails to load on older GPUs when `CUDA_FORCE_PTX_JIT=1` is set. (See [Application Compatibility](https:\/\/docs.nvidia.com\/cuda\/cuda-c-programming-guide\/index.html#application-compatibility) for details.)\n## System requirements\n- Ubuntu 16.04 or higher (64-bit)\n- macOS 12.0 (Monterey) or higher (64-bit) *(no GPU support)*\n- Windows Native - Windows 7 or higher (64-bit) *(no GPU support after TF 2.10)*\n- Windows WSL2 - Windows 10 19044 or higher (64-bit)\n## Software requirements\n- Python 3.9–3.12\n- pip version 19.0 or higher for Linux (requires `manylinux2014` support) and Windows. pip version 20.3 or higher for macOS.\n- Windows Native Requires [Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019](https:\/\/learn.microsoft.com\/en-us\/cpp\/windows\/latest-supported-vc-redist)\n\nThe following NVIDIA® software are only required for GPU support.\n\n- [NVIDIA® GPU drivers](https:\/\/www.nvidia.com\/drivers)\n  - \\>= 525.60.13 for Linux\n  - \\>= 528.33 for WSL on Windows\n- [CUDA® Toolkit 12.3](https:\/\/developer.nvidia.com\/cuda-toolkit-archive).\n- [cuDNN SDK 8.9.7](https:\/\/developer.nvidia.com\/cudnn).\n- *(Optional)* [TensorRT](https:\/\/docs.nvidia.com\/deeplearning\/tensorrt\/archives\/index.html#trt_7) to improve latency and throughput for inference.\n## Step-by-step instructions\n### 1\\. System requirements\n- Ubuntu 16.04 or higher (64-bit)\n\nTensorFlow only officially supports Ubuntu. However, the following instructions may also work for other Linux distros.\n\n### 2\\. GPU setup\nYou can skip this section if you only run TensorFlow on the CPU.\n\nInstall the [NVIDIA GPU driver](https:\/\/www.nvidia.com\/Download\/index.aspx) if you have not. You can use the following command to verify it is installed.\n```\nnvidia-smi\n```\n### 3\\. Create a virtual environment with [venv](https:\/\/docs.python.org\/3\/library\/venv.html)\nThe venv module is part of Python’s standard library and is the officially recommended way to create virtual environments.\n\nNavigate to your desired virtual environments directory and create a new venv environment named [`tf`](https:\/\/www.tensorflow.org\/api_docs\/python\/tf) with the following command.\n```\npython3 -m venv tf\n```\nYou can activate it with the following command.\n```\nsource tf\/bin\/activate\n```\nMake sure that the virtual environment is activated for the rest of the installation.\n\n### 4\\. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.\n```\npip install --upgrade pip\n```\nThen, install TensorFlow with pip.\n```\n# For GPU users\npip install tensorflow[and-cuda]\n# For CPU users\npip install tensorflow\n```\n### 6\\. Verify the installation\nVerify the CPU setup:\n```\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\nIf a tensor is returned, you've installed TensorFlow successfully.\n\nVerify the GPU setup:\n```\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\nIf a list of GPU devices is returned, you've installed TensorFlow successfully. **If not continue to the next step**.\n\n### 6\\. \\[GPU only\\] Virtual environment configuration\nIf the GPU test in the last section was unsuccessful, the most likely cause is that components aren't being detected, and\/or conflict with the existing system CUDA installation. So you need to add some symbolic links to fix this.\n\n- Create symbolic links to NVIDIA shared libraries:\n```\npushd $(dirname $(python -c 'print(__import__(\"tensorflow\").__file__)'))\nln -svf ..\/nvidia\/*\/lib\/*.so* .\npopd\n```\n- Create a symbolic link to ptxas:\n```\nln -sf $(find $(dirname $(dirname $(python -c \"import nvidia.cuda_nvcc;         \nprint(nvidia.cuda_nvcc.__file__)\"))\/*\/bin\/) -name ptxas -print -quit) $VIRTUAL_ENV\/bin\/ptxas\n```\nVerify the GPU setup:\n```\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\n\n### 1\\. System requirements\n- macOS 10.12.6 (Sierra) or higher (64-bit)\n\nCurrently there is no official GPU support for running TensorFlow on MacOS. The following instructions are for running on CPU.\n\n### 2\\. Check Python version\nCheck if your Python environment is already configured:\n```\npython3 --version\npython3 -m pip --version\n```\n### 3\\. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.\n```\npip install --upgrade pip\n```\nThen, install TensorFlow with pip.\n```\npip install tensorflow\n```\n### 4\\. Verify the installation\n```\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\nIf a tensor is returned, you've installed TensorFlow successfully.\n\n## 1\\. System requirements\n- Windows 7 or higher (64-bit)\n### 2\\. Install Microsoft Visual C++ Redistributable\nInstall the *Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017, and 2019*. Starting with the TensorFlow 2.1.0 version, the `msvcp140_1.dll` file is required from this package (which may not be provided from older redistributable packages). The redistributable comes with *Visual Studio 2019* but can be installed separately:\n\n1. Go to the [Microsoft Visual C++ downloads](https:\/\/support.microsoft.com\/help\/2977003\/the-latest-supported-visual-c-downloads).\n2. Scroll down the page to the *Visual Studio 2015, 2017 and 2019* section.\n3. Download and install the *Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019* for your platform.\n\nMake sure [long paths are enabled](https:\/\/superuser.com\/questions\/1119883\/windows-10-enable-ntfs-long-paths-policy-option-missing) on Windows.\n\n### 3\\. Install Miniconda\n[Miniconda](https:\/\/docs.conda.io\/en\/latest\/miniconda.html) is the recommended approach for installing TensorFlow with GPU support. It creates a separate environment to avoid changing any installed software in your system. This is also the easiest way to install the required software especially for the GPU setup.\n\nDownload the [Miniconda Windows Installer](https:\/\/repo.anaconda.com\/miniconda\/Miniconda3-latest-Windows-x86_64.exe). Double-click the downloaded file and follow the instructions on the screen.\n\n### 4\\. Create a conda environment\nCreate a new conda environment named [`tf`](https:\/\/www.tensorflow.org\/api_docs\/python\/tf) with the following command.\n```\nconda create --name tf python=3.9\n```\nYou can deactivate and activate it with the following commands.\n```\nconda deactivate\nconda activate tf\n```\nMake sure it is activated for the rest of the installation.\n\n### 5\\. GPU setup\nYou can skip this section if you only run TensorFlow on CPU.\n\nFirst install [NVIDIA GPU driver](https:\/\/www.nvidia.com\/Download\/index.aspx) if you have not.\n\nThen install the CUDA, cuDNN with conda.\n```\nconda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0\n```\n### 6\\. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.\n```\npip install --upgrade pip\n```\nThen, install TensorFlow with pip.\n```\n# Anything above 2.10 is not supported on the GPU on Windows Native\npip install \"tensorflow<2.11\"\n```\n### 7\\. Verify the installation\nVerify the CPU setup:\n```\npython -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\nIf a tensor is returned, you've installed TensorFlow successfully.\n\nVerify the GPU setup:\n```\npython -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\nIf a list of GPU devices is returned, you've installed TensorFlow successfully.\n\n### 1\\. System requirements\n- Windows 10 19044 or higher (64-bit). This corresponds to Windows 10 version 21H2, the November 2021 update.\n\nSee the following documents to:\n\n- [Download the latest Windows 10 update](https:\/\/www.microsoft.com\/software-download\/windows10).\n- [Install WSL2](https:\/\/docs.microsoft.com\/windows\/wsl\/install)\n- [Setup NVIDIA® GPU support in WSL2](https:\/\/docs.nvidia.com\/cuda\/wsl-user-guide\/index.html)\n### 2\\. GPU setup\nYou can skip this section if you only run TensorFlow on the CPU.\n\nInstall the [NVIDIA GPU driver](https:\/\/www.nvidia.com\/Download\/index.aspx) if you have not. You can use the following command to verify it is installed.\n```\nnvidia-smi\n```\n### 3\\. Install TensorFlow\nTensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.\n```\npip install --upgrade pip\n```\nThen, install TensorFlow with pip.\n```\n# For GPU users\npip install tensorflow[and-cuda]\n# For CPU users\npip install tensorflow\n```\n### 4\\. Verify the installation\nVerify the CPU setup:\n```\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\n```\nIf a tensor is returned, you've installed TensorFlow successfully.\n\nVerify the GPU setup:\n```\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\n```\nIf a list of GPU devices is returned, you've installed TensorFlow successfully.\n## Package location\nA few installation mechanisms require the URL of the TensorFlow Python package. The value you specify depends on your Python version.\n\n## Python version support\n| Version | URL |\n|---|---|\n| Linux x86 |   |\n| Python 3.10 GPU support | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl) |\n| Python 3.10 CPU-only | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow\\_cpu-2.21.0-cp310-cp310-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow_cpu-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl) |\n| Python 3.11 GPU support | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl) |\n| Python 3.11 CPU-only | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow\\_cpu-2.21.0-cp311-cp311-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow_cpu-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl) |\n| Python 3.12 GPU support | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_x86_64.whl) |\n| Python 3.12 CPU-only | <https:\/\/storage.googleapis.com\/tensorflow\/versions\/2.20.0\/tensorflow_cpu-2.20.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl> |\n| Python 3.13 GPU support | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl) |\n| Python 3.13 CPU-only | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow\\_cpu-2.21.0-cp313-cp313-manylinux\\_2\\_27\\_x86\\_64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow_cpu-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl) |\n| Linux Arm64 (CPU-only) |   |\n| Python 3.10 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-manylinux\\_2\\_27\\_aarch64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_aarch64.whl) |\n| Python 3.11 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-manylinux\\_2\\_27\\_aarch64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_aarch64.whl) |\n| Python 3.12 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-manylinux\\_2\\_27\\_aarch64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_aarch64.whl) |\n| Python 3.13 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-manylinux\\_2\\_27\\_aarch64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_aarch64.whl) |\n| macOS x86 (CPU-only) |   |\n| **Caution**: TensorFlow 2.16 was the **last** TensorFlow release that supported macOS x86 |   |\n| Python 3.10 | <https:\/\/storage.googleapis.com\/tensorflow\/versions\/2.16.2\/tensorflow-2.16.2-cp310-cp310-macosx_10_15_x86_64.whl> |\n| Python 3.11 | <https:\/\/storage.googleapis.com\/tensorflow\/versions\/2.16.2\/tensorflow-2.16.2-cp311-cp311-macosx_10_15_x86_64.whl> |\n| Python 3.12 | <https:\/\/storage.googleapis.com\/tensorflow\/versions\/2.16.2\/tensorflow-2.16.2-cp312-cp312-macosx_10_15_x86_64.whl> |\n| macOS Arm64 (CPU-only) |   |\n| Python 3.10 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-macosx\\_12\\_0\\_arm64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-macosx_12_0_arm64.whl) |\n| Python 3.11 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-macosx\\_12\\_0\\_arm64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-macosx_12_0_arm64.whl) |\n| Python 3.12 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-macosx\\_12\\_0\\_arm64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-macosx_12_0_arm64.whl) |\n| Python 3.13 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-macosx\\_12\\_0\\_arm64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-macosx_12_0_arm64.whl) |\n| Windows (CPU-only) |   |\n| Python 3.10 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-win\\_amd64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp310-cp310-win_amd64.whl) |\n| Python 3.11 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-win\\_amd64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp311-cp311-win_amd64.whl) |\n| Python 3.12 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-win\\_amd64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp312-cp312-win_amd64.whl) |\n| Python 3.13 | [https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-win\\_amd64.whl](https:\/\/storage.googleapis.com\/tensorflow\/versions\/%222.21.0%22\/tensorflow-2.21.0-cp313-cp313-win_amd64.whl) |\n\nExcept as otherwise noted, the content of this page is licensed under the [Creative Commons Attribution 4.0 License](https:\/\/creativecommons.org\/licenses\/by\/4.0\/), and code samples are licensed under the [Apache 2.0 License](https:\/\/www.apache.org\/licenses\/LICENSE-2.0). For details, see the [Google Developers Site Policies](https:\/\/developers.google.com\/site-policies). Java is a registered trademark of Oracle and\/or its affiliates.\n\nLast updated 2026-03-12 UTC.","meta_canonical":null}

3. Robots.txt Check

Query:

Response:

4. Spam/Ban Check

Query:

Response:

5. Seen Status Check

ℹ️ Skipped - page is already crawled

📄

INDEXABLE

✅

CRAWLED

5 hours ago

🤖

ROBOTS ALLOWED

Page Info Filters

Filter	Status	Condition	Details
HTTP status	PASS	`download_http_code = 200`	HTTP 200
Age cutoff	PASS	`download_stamp > now() - 6 MONTH`	0 months ago
History drop	PASS	`isNull(history_drop_reason)`	No drop reason
Spam/ban	PASS	`fh_dont_index != 1 AND ml_spam_score = 0`	ml_spam_score=0
Canonical	PASS	`meta_canonical IS NULL OR = '' OR = src_unparsed`	Not set

Page Details

Property	Value
URL	https://www.tensorflow.org/install/pip
Last Crawled	2026-04-10 05:34:02 (5 hours ago)
First Indexed	2018-09-22 21:01:52 (7 years ago)
HTTP Status Code	200
Meta Title	Install TensorFlow with pip
Meta Description	null
Meta Canonical	null
Boilerpipe Text	Install Tensor Flow with pip Stay organized with collections Save and categorize content based on your preferences. On this page Hardware requirements System requirements Software requirements Step-by-step instructions Package location Python version support This guide is for the latest stable version of TensorFlow. For the preview build (nightly) , use the pip package named tf-nightly . Refer to these tables for older TensorFlow version requirements. For the CPU-only build, use the pip package named tensorflow-cpu . Here are the quick versions of the install commands. Scroll down for the step-by-step instructions. python3 -m pip install 'tensorflow[and-cuda]' # Verify the installation: python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" # There is currently no official GPU support for MacOS. python3 -m pip install tensorflow # Verify the installation: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" conda install -c conda-forge cudatoolkit = 11 .2 cudnn = 8 .1.0 # Anything above 2.10 is not supported on the GPU on Windows Native python -m pip install "tensorflow<2.11" # Verify the installation: python -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" python3 -m pip install tensorflow [ and-cuda ] # Verify the installation: python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" python3 -m pip install tensorflow # Verify the installation: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" python3 -m pip install tf-nightly # Verify the installation: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" Hardware requirements The following GPU-enabled devices are supported: NVIDIA® GPU card with CUDA® architectures 3.5, 5.0, 6.0, 7.0, 7.5, 8.0 and higher. See the list of CUDA®-enabled GPU cards . For GPUs with unsupported CUDA® architectures, or to avoid JIT compilation from PTX, or to use different versions of the NVIDIA® libraries, see the Linux build from source guide. Packages do not contain PTX code except for the latest supported CUDA® architecture; therefore, TensorFlow fails to load on older GPUs when CUDA_FORCE_PTX_JIT=1 is set. (See Application Compatibility for details.) System requirements Ubuntu 16.04 or higher (64-bit) macOS 12.0 (Monterey) or higher (64-bit) (no GPU support) Windows Native - Windows 7 or higher (64-bit) (no GPU support after TF 2.10) Windows WSL2 - Windows 10 19044 or higher (64-bit) Software requirements Python 3.9–3.12 pip version 19.0 or higher for Linux (requires manylinux2014 support) and Windows. pip version 20.3 or higher for macOS. Windows Native Requires Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019 The following NVIDIA® software are only required for GPU support. NVIDIA® GPU drivers >= 525.60.13 for Linux >= 528.33 for WSL on Windows CUDA® Toolkit 12.3 . cuDNN SDK 8.9.7 . (Optional) TensorRT to improve latency and throughput for inference. Step-by-step instructions 1. System requirements Ubuntu 16.04 or higher (64-bit) TensorFlow only officially supports Ubuntu. However, the following instructions may also work for other Linux distros. 2. GPU setup You can skip this section if you only run TensorFlow on the CPU. Install the NVIDIA GPU driver if you have not. You can use the following command to verify it is installed. nvidia-smi 3. Create a virtual environment with venv The venv module is part of Python’s standard library and is the officially recommended way to create virtual environments. Navigate to your desired virtual environments directory and create a new venv environment named tf with the following command. python3 -m venv tf You can activate it with the following command. source tf/bin/activate Make sure that the virtual environment is activated for the rest of the installation. 4. Install Tensor Flow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. pip install --upgrade pip Then, install TensorFlow with pip. # For GPU users pip install tensorflow [ and-cuda ] # For CPU users pip install tensorflow 6. Verify the installation Verify the CPU setup: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" If a tensor is returned, you've installed TensorFlow successfully. Verify the GPU setup: python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" If a list of GPU devices is returned, you've installed TensorFlow successfully. If not continue to the next step . 6. [GPU only] Virtual environment configuration If the GPU test in the last section was unsuccessful, the most likely cause is that components aren't being detected, and/or conflict with the existing system CUDA installation. So you need to add some symbolic links to fix this. Create symbolic links to NVIDIA shared libraries: pushd $( dirname $( python -c 'print(__import__("tensorflow").__file__)' )) ln -svf ../nvidia//lib/.so* . popd Create a symbolic link to ptxas: ln -sf $( find $( dirname $( dirname $( python -c "import nvidia.cuda_nvcc; print(nvidia.cuda_nvcc.__file__)" )) /*/bin/ ) -name ptxas -print -quit ) $VIRTUAL_ENV /bin/ptxas Verify the GPU setup: python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" 1. System requirements macOS 10.12.6 (Sierra) or higher (64-bit) Currently there is no official GPU support for running TensorFlow on MacOS. The following instructions are for running on CPU. 2. Check Python version Check if your Python environment is already configured: python3 --version python3 -m pip --version 3. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. pip install --upgrade pip Then, install TensorFlow with pip. pip install tensorflow 4. Verify the installation python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" If a tensor is returned, you've installed TensorFlow successfully. 1. System requirements Windows 7 or higher (64-bit) 2. Install Microsoft Visual C++ Redistributable Install the Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017, and 2019 . Starting with the TensorFlow 2.1.0 version, the msvcp140_1.dll file is required from this package (which may not be provided from older redistributable packages). The redistributable comes with Visual Studio 2019 but can be installed separately: Go to the Microsoft Visual C++ downloads . Scroll down the page to the Visual Studio 2015, 2017 and 2019 section. Download and install the Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019 for your platform. Make sure long paths are enabled on Windows. 3. Install Miniconda Miniconda is the recommended approach for installing TensorFlow with GPU support. It creates a separate environment to avoid changing any installed software in your system. This is also the easiest way to install the required software especially for the GPU setup. Download the Miniconda Windows Installer . Double-click the downloaded file and follow the instructions on the screen. 4. Create a conda environment Create a new conda environment named tf with the following command. conda create --name tf python = 3 .9 You can deactivate and activate it with the following commands. conda deactivate conda activate tf Make sure it is activated for the rest of the installation. 5. GPU setup You can skip this section if you only run TensorFlow on CPU. First install NVIDIA GPU driver if you have not. Then install the CUDA, cuDNN with conda. conda install -c conda-forge cudatoolkit = 11 .2 cudnn = 8 .1.0 6. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. pip install --upgrade pip Then, install TensorFlow with pip. # Anything above 2.10 is not supported on the GPU on Windows Native pip install "tensorflow<2.11" 7. Verify the installation Verify the CPU setup: python -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" If a tensor is returned, you've installed TensorFlow successfully. Verify the GPU setup: python -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" If a list of GPU devices is returned, you've installed TensorFlow successfully. 1. System requirements Windows 10 19044 or higher (64-bit). This corresponds to Windows 10 version 21H2, the November 2021 update. See the following documents to: Download the latest Windows 10 update . Install WSL2 Setup NVIDIA® GPU support in WSL2 2. GPU setup You can skip this section if you only run TensorFlow on the CPU. Install the NVIDIA GPU driver if you have not. You can use the following command to verify it is installed. nvidia-smi 3. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. pip install --upgrade pip Then, install TensorFlow with pip. # For GPU users pip install tensorflow [ and-cuda ] # For CPU users pip install tensorflow 4. Verify the installation Verify the CPU setup: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" If a tensor is returned, you've installed TensorFlow successfully. Verify the GPU setup: python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" If a list of GPU devices is returned, you've installed TensorFlow successfully. Package location A few installation mechanisms require the URL of the TensorFlow Python package. The value you specify depends on your Python version. Python version support Version URL Linux x86 Python 3.10 GPU support https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl Python 3.10 CPU-only https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow_cpu-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl Python 3.11 GPU support https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl Python 3.11 CPU-only https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow_cpu-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl Python 3.12 GPU support https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_x86_64.whl Python 3.12 CPU-only https://storage.googleapis.com/tensorflow/versions/2.20.0/tensorflow_cpu-2.20.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl Python 3.13 GPU support https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl Python 3.13 CPU-only https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow_cpu-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl Linux Arm64 (CPU-only) Python 3.10 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_aarch64.whl Python 3.11 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_aarch64.whl Python 3.12 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_aarch64.whl Python 3.13 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_aarch64.whl macOS x86 (CPU-only) Caution : TensorFlow 2.16 was the last TensorFlow release that supported macOS x86 Python 3.10 https://storage.googleapis.com/tensorflow/versions/2.16.2/tensorflow-2.16.2-cp310-cp310-macosx_10_15_x86_64.whl Python 3.11 https://storage.googleapis.com/tensorflow/versions/2.16.2/tensorflow-2.16.2-cp311-cp311-macosx_10_15_x86_64.whl Python 3.12 https://storage.googleapis.com/tensorflow/versions/2.16.2/tensorflow-2.16.2-cp312-cp312-macosx_10_15_x86_64.whl macOS Arm64 (CPU-only) Python 3.10 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-macosx_12_0_arm64.whl Python 3.11 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-macosx_12_0_arm64.whl Python 3.12 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-macosx_12_0_arm64.whl Python 3.13 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-macosx_12_0_arm64.whl Windows (CPU-only) Python 3.10 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-win_amd64.whl Python 3.11 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-win_amd64.whl Python 3.12 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-win_amd64.whl Python 3.13 https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-win_amd64.whl Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License , and code samples are licensed under the Apache 2.0 License . For details, see the Google Developers Site Policies . Java is a registered trademark of Oracle and/or its affiliates. Last updated 2026-03-12 UTC. [[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2026-03-12 UTC."],[],[]]
Markdown	[Skip to main content](https://www.tensorflow.org/install/pip#main-content) [![TensorFlow](https://www.gstatic.com/devrel-devsite/prod/v369eac9380f92e8fedc492e2689927bb3475d758266c381eee326d0b49a77481/tensorflow/images/lockup.svg)](https://www.tensorflow.org/) [Install](https://www.tensorflow.org/install) [Learn](https://www.tensorflow.org/learn) - [Introduction New to TensorFlow?](https://www.tensorflow.org/learn) - [Tutorials Learn how to use TensorFlow with end-to-end examples](https://www.tensorflow.org/tutorials) - [Guide Learn framework concepts and components](https://www.tensorflow.org/guide) - [Learn ML Educational resources to master your path with TensorFlow](https://www.tensorflow.org/resources/learn-ml) More [API](https://www.tensorflow.org/api) - [TensorFlow (v2.16.1)](https://www.tensorflow.org/api/stable) - [Versions…](https://www.tensorflow.org/versions) - [TensorFlow.js](https://js.tensorflow.org/api/latest/) - [TensorFlow Lite](https://www.tensorflow.org/lite/api_docs) - [TFX](https://www.tensorflow.org/tfx/api_docs) [Ecosystem](https://www.tensorflow.org/resources/models-datasets) - LIBRARIES - [TensorFlow.js Develop web ML applications in JavaScript](https://www.tensorflow.org/js) - [TensorFlow Lite Deploy ML on mobile, microcontrollers and other edge devices](https://www.tensorflow.org/lite) - [TFX Build production ML pipelines](https://www.tensorflow.org/tfx) - [All libraries Create advanced models and extend TensorFlow](https://www.tensorflow.org/resources/libraries-extensions) - RESOURCES - [Models & datasets Pre-trained models and datasets built by Google and the community](https://www.tensorflow.org/resources/models-datasets) - [Tools Tools to support and accelerate TensorFlow workflows](https://www.tensorflow.org/resources/tools) - [Responsible AI Resources for every stage of the ML workflow](https://www.tensorflow.org/responsible_ai) - [Recommendation systems Build recommendation systems with open source tools](https://www.tensorflow.org/resources/recommendation-systems) [Community](https://www.tensorflow.org/community) - [Groups User groups, interest groups and mailing lists](https://www.tensorflow.org/community/groups) - [Contribute Guide for contributing to code and documentation](https://www.tensorflow.org/community/contribute) - [Blog Stay up to date with all things TensorFlow](https://blog.tensorflow.org/) - [Forum Discussion platform for the TensorFlow community](https://discuss.tensorflow.org/) [Why TensorFlow](https://www.tensorflow.org/about) - [About](https://www.tensorflow.org/about) - [Case studies](https://www.tensorflow.org/about/case-studies) - [English](https://www.tensorflow.org/install/pip) - [Español](https://www.tensorflow.org/install/pip?hl=es) - [Español – América Latina](https://www.tensorflow.org/install/pip?hl=es-419) - [Français](https://www.tensorflow.org/install/pip?hl=fr) - [Indonesia](https://www.tensorflow.org/install/pip?hl=id) - [Italiano](https://www.tensorflow.org/install/pip?hl=it) - [Polski](https://www.tensorflow.org/install/pip?hl=pl) - [Português](https://www.tensorflow.org/install/pip?hl=pt) - [Português – Brasil](https://www.tensorflow.org/install/pip?hl=pt-br) - [Tiếng Việt](https://www.tensorflow.org/install/pip?hl=vi) - [Türkçe](https://www.tensorflow.org/install/pip?hl=tr) - [Русский](https://www.tensorflow.org/install/pip?hl=ru) - [עברית](https://www.tensorflow.org/install/pip?hl=he) - [العربيّة](https://www.tensorflow.org/install/pip?hl=ar) - [فارسی](https://www.tensorflow.org/install/pip?hl=fa) - [हिंदी](https://www.tensorflow.org/install/pip?hl=hi) - [বাংলা](https://www.tensorflow.org/install/pip?hl=bn) - [ภาษาไทย](https://www.tensorflow.org/install/pip?hl=th) - [中文 – 简体](https://www.tensorflow.org/install/pip?hl=zh-cn) - [中文 – 繁體](https://www.tensorflow.org/install/pip?hl=zh-tw) - [日本語](https://www.tensorflow.org/install/pip?hl=ja) - [한국어](https://www.tensorflow.org/install/pip?hl=ko) [GitHub](https://github.com/tensorflow) [Sign in](https://www.tensorflow.org/_d/signin?continue=https%3A%2F%2Fwww.tensorflow.org%2Finstall%2Fpip&prompt=select_account) - [Install](https://www.tensorflow.org/install) [![TensorFlow](https://www.gstatic.com/devrel-devsite/prod/v369eac9380f92e8fedc492e2689927bb3475d758266c381eee326d0b49a77481/tensorflow/images/lockup.svg)](https://www.tensorflow.org/) - [Install](https://www.tensorflow.org/install) - [Learn](https://www.tensorflow.org/learn) - More - [API](https://www.tensorflow.org/api) - More - [Ecosystem](https://www.tensorflow.org/resources/models-datasets) - More - [Community](https://www.tensorflow.org/community) - More - [Why TensorFlow](https://www.tensorflow.org/about) - More - [GitHub](https://github.com/tensorflow) - [Install Tensor Flow](https://www.tensorflow.org/install) - Packages - [pip](https://www.tensorflow.org/install/pip) - [Docker](https://www.tensorflow.org/install/docker) - Additional setup - [GPU device plugins](https://www.tensorflow.org/install/gpu_plugins) - [Problems](https://www.tensorflow.org/install/errors) - Build from source - [Linux / mac OS](https://www.tensorflow.org/install/source) - [Windows](https://www.tensorflow.org/install/source_windows) - [SIG Build](https://github.com/tensorflow/build) - Language bindings - [Java](https://www.tensorflow.org/jvm/install) - [Java (legacy)](https://www.tensorflow.org/install/lang_java_legacy) - [C](https://www.tensorflow.org/install/lang_c) - [Go](https://github.com/tensorflow/build/tree/master/golang_install_guide) - [Introduction](https://www.tensorflow.org/learn) - [Tutorials](https://www.tensorflow.org/tutorials) - [Guide](https://www.tensorflow.org/guide) - [Learn ML](https://www.tensorflow.org/resources/learn-ml) - [TensorFlow (v2.16.1)](https://www.tensorflow.org/api/stable) - [Versions…](https://www.tensorflow.org/versions) - [TensorFlow.js](https://js.tensorflow.org/api/latest/) - [TensorFlow Lite](https://www.tensorflow.org/lite/api_docs) - [TFX](https://www.tensorflow.org/tfx/api_docs) - LIBRARIES - [TensorFlow.js](https://www.tensorflow.org/js) - [TensorFlow Lite](https://www.tensorflow.org/lite) - [TFX](https://www.tensorflow.org/tfx) - [All libraries](https://www.tensorflow.org/resources/libraries-extensions) - RESOURCES - [Models & datasets](https://www.tensorflow.org/resources/models-datasets) - [Tools](https://www.tensorflow.org/resources/tools) - [Responsible AI](https://www.tensorflow.org/responsible_ai) - [Recommendation systems](https://www.tensorflow.org/resources/recommendation-systems) - [Groups](https://www.tensorflow.org/community/groups) - [Contribute](https://www.tensorflow.org/community/contribute) - [Blog](https://blog.tensorflow.org/) - [Forum](https://discuss.tensorflow.org/) - [About](https://www.tensorflow.org/about) - [Case studies](https://www.tensorflow.org/about/case-studies) - On this page - [Hardware requirements](https://www.tensorflow.org/install/pip#hardware_requirements) - [System requirements](https://www.tensorflow.org/install/pip#system_requirements) - [Software requirements](https://www.tensorflow.org/install/pip#software_requirements) - [Step-by-step instructions](https://www.tensorflow.org/install/pip#step-by-step_instructions) - [Package location](https://www.tensorflow.org/install/pip#package_location) - [Python version support](https://www.tensorflow.org/install/pip#python_version_support) - [TensorFlow](https://www.tensorflow.org/) - [Install](https://www.tensorflow.org/install) Was this helpful? # Install TensorFlow with pipStay organized with collections Save and categorize content based on your preferences. - On this page - [Hardware requirements](https://www.tensorflow.org/install/pip#hardware_requirements) - [System requirements](https://www.tensorflow.org/install/pip#system_requirements) - [Software requirements](https://www.tensorflow.org/install/pip#software_requirements) - [Step-by-step instructions](https://www.tensorflow.org/install/pip#step-by-step_instructions) - [Package location](https://www.tensorflow.org/install/pip#package_location) - [Python version support](https://www.tensorflow.org/install/pip#python_version_support) This guide is for the latest stable version of TensorFlow. For the preview build (nightly), use the pip package named `tf-nightly`. Refer to [these tables](https://www.tensorflow.org/install/source#tested_build_configurations) for older TensorFlow version requirements. For the CPU-only build, use the pip package named `tensorflow-cpu`. Here are the quick versions of the install commands. Scroll down for the step-by-step instructions. [Linux](https://www.tensorflow.org/install/pip#linux) [MacOS](https://www.tensorflow.org/install/pip#macos) [Windows Native](https://www.tensorflow.org/install/pip#windows-native) [Windows WSL2](https://www.tensorflow.org/install/pip#windows-wsl2) [CPU](https://www.tensorflow.org/install/pip#cpu) [Nightly](https://www.tensorflow.org/install/pip#nightly) More Note: Starting with TensorFlow `2.10`, Linux CPU-builds for Aarch64/ARM64 processors are built, maintained, tested and released by a third party: [AWS](https://aws.amazon.com/). Installing the [`tensorflow`](https://pypi.org/project/tensorflow/) package on an ARM machine installs AWS's [`tensorflow-cpu-aws`](https://pypi.org/project/tensorflow-cpu-aws/) package. They are provided as-is. Tensorflow will use reasonable efforts to maintain the availability and integrity of this pip package. There may be delays if the third party fails to release the pip package. See [this blog post](https://blog.tensorflow.org/2022/09/announcing-tensorflow-official-build-collaborators.html) for more information about this collaboration. ``` python3 -m pip install 'tensorflow[and-cuda]' # Verify the installation: python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` ``` # There is currently no official GPU support for MacOS. python3 -m pip install tensorflow # Verify the installation: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` Caution: TensorFlow `2.10` was the last TensorFlow release that supported GPU on native-Windows. Starting with TensorFlow `2.11`, you will need to install [TensorFlow in WSL2](https://tensorflow.org/install/pip#windows-wsl2), or install `tensorflow` or `tensorflow-cpu` and, optionally, try the [TensorFlow-DirectML-Plugin](https://github.com/microsoft/tensorflow-directml-plugin#tensorflow-directml-plugin-) ``` conda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0 # Anything above 2.10 is not supported on the GPU on Windows Native python -m pip install "tensorflow<2.11" # Verify the installation: python -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` Note: TensorFlow with GPU access is supported for WSL2 on Windows 10 19044 or higher. This corresponds to Windows 10 version 21H2, the November 2021 update. You can get the latest update from here: [Download Windows 10](https://www.microsoft.com/software-download/windows10). For instructions, see [Install WSL2](https://docs.microsoft.com/windows/wsl/install) and [NVIDIA’s setup docs](https://docs.nvidia.com/cuda/wsl-user-guide/index.html) for CUDA in WSL. ``` python3 -m pip install tensorflow[and-cuda] # Verify the installation: python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` Note: Starting with TensorFlow `2.10`, Windows CPU-builds for x86/x64 processors are built, maintained, tested and released by a third party: [Intel](https://www.intel.com/). Installing the Windows-native [`tensorflow`](https://pypi.org/project/tensorflow/) or [`tensorflow-cpu`](https://pypi.org/project/tensorflow-cpu/) package installs Intel's [`tensorflow-intel`](https://pypi.org/project/tensorflow-intel/) package. These packages are provided as-is. Tensorflow will use reasonable efforts to maintain the availability and integrity of this pip package. There may be delays if the third party fails to release the pip package. See [this blog post](https://blog.tensorflow.org/2022/09/announcing-tensorflow-official-build-collaborators.html) for more information about this collaboration. ``` python3 -m pip install tensorflow # Verify the installation: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` ``` python3 -m pip install tf-nightly # Verify the installation: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` ## Hardware requirements Note: TensorFlow binaries use [AVX instructions](https://en.wikipedia.org/wiki/Advanced_Vector_Extensions#CPUs_with_AVX) which may not run on older CPUs. The following GPU-enabled devices are supported: - NVIDIA® GPU card with CUDA® architectures 3.5, 5.0, 6.0, 7.0, 7.5, 8.0 and higher. See the list of [CUDA®-enabled GPU cards](https://developer.nvidia.com/cuda-gpus). - For GPUs with unsupported CUDA® architectures, or to avoid JIT compilation from PTX, or to use different versions of the NVIDIA® libraries, see the [Linux build from source](https://www.tensorflow.org/install/source) guide. - Packages do not contain PTX code except for the latest supported CUDA® architecture; therefore, TensorFlow fails to load on older GPUs when `CUDA_FORCE_PTX_JIT=1` is set. (See [Application Compatibility](https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#application-compatibility) for details.) Note: The error message "Status: device kernel image is invalid" indicates that the TensorFlow package does not contain PTX for your architecture. You can enable compute capabilities by [building TensorFlow from source](https://www.tensorflow.org/install/source). ## System requirements - Ubuntu 16.04 or higher (64-bit) - macOS 12.0 (Monterey) or higher (64-bit) (no GPU support) - Windows Native - Windows 7 or higher (64-bit) (no GPU support after TF 2.10) - Windows WSL2 - Windows 10 19044 or higher (64-bit) Note: GPU support is available for Ubuntu and Windows with CUDA®-enabled cards. ## Software requirements - Python 3.9–3.12 - pip version 19.0 or higher for Linux (requires `manylinux2014` support) and Windows. pip version 20.3 or higher for macOS. - Windows Native Requires [Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019](https://learn.microsoft.com/en-us/cpp/windows/latest-supported-vc-redist) The following NVIDIA® software are only required for GPU support. - [NVIDIA® GPU drivers](https://www.nvidia.com/drivers) - \>= 525.60.13 for Linux - \>= 528.33 for WSL on Windows - [CUDA® Toolkit 12.3](https://developer.nvidia.com/cuda-toolkit-archive). - [cuDNN SDK 8.9.7](https://developer.nvidia.com/cudnn). - (Optional) [TensorRT](https://docs.nvidia.com/deeplearning/tensorrt/archives/index.html#trt_7) to improve latency and throughput for inference. ## Step-by-step instructions [Linux](https://www.tensorflow.org/install/pip#linux) [MacOS](https://www.tensorflow.org/install/pip#macos) [Windows Native](https://www.tensorflow.org/install/pip#windows-native) [Windows WSL2](https://www.tensorflow.org/install/pip#windows-wsl2) More ### 1\. System requirements - Ubuntu 16.04 or higher (64-bit) TensorFlow only officially supports Ubuntu. However, the following instructions may also work for other Linux distros. Note: Starting with TensorFlow `2.10`, Linux CPU-builds for Aarch64/ARM64 processors are built, maintained, tested and released by a third party: [AWS](https://aws.amazon.com/). Installing the [`tensorflow`](https://pypi.org/project/tensorflow/) package on an ARM machine installs AWS's [`tensorflow-cpu-aws`](https://pypi.org/project/tensorflow-cpu-aws/) package. They are provided as-is. Tensorflow will use reasonable efforts to maintain the availability and integrity of this pip package. There may be delays if the third party fails to release the pip package. See [this blog post](https://blog.tensorflow.org/2022/09/announcing-tensorflow-official-build-collaborators.html) for more information about this collaboration. ### 2\. GPU setup You can skip this section if you only run TensorFlow on the CPU. Install the [NVIDIA GPU driver](https://www.nvidia.com/Download/index.aspx) if you have not. You can use the following command to verify it is installed. ``` nvidia-smi ``` ### 3\. Create a virtual environment with [venv](https://docs.python.org/3/library/venv.html) The venv module is part of Python’s standard library and is the officially recommended way to create virtual environments. Navigate to your desired virtual environments directory and create a new venv environment named [`tf`](https://www.tensorflow.org/api_docs/python/tf) with the following command. ``` python3 -m venv tf ``` You can activate it with the following command. ``` source tf/bin/activate ``` Make sure that the virtual environment is activated for the rest of the installation. ### 4\. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. ``` pip install --upgrade pip ``` Then, install TensorFlow with pip. ``` # For GPU users pip install tensorflow[and-cuda] # For CPU users pip install tensorflow ``` Note: Do not install TensorFlow with `conda`. It may not have the latest stable version. `pip` is recommended since TensorFlow is only officially released to PyPI. ### 6\. Verify the installation Verify the CPU setup: ``` python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` If a tensor is returned, you've installed TensorFlow successfully. Verify the GPU setup: ``` python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` If a list of GPU devices is returned, you've installed TensorFlow successfully. If not continue to the next step. ### 6\. \[GPU only\] Virtual environment configuration If the GPU test in the last section was unsuccessful, the most likely cause is that components aren't being detected, and/or conflict with the existing system CUDA installation. So you need to add some symbolic links to fix this. - Create symbolic links to NVIDIA shared libraries: ``` pushd $(dirname $(python -c 'print(__import__("tensorflow").__file__)')) ln -svf ../nvidia//lib/.so* . popd ``` - Create a symbolic link to ptxas: ``` ln -sf $(find $(dirname $(dirname $(python -c "import nvidia.cuda_nvcc; print(nvidia.cuda_nvcc.__file__)"))//bin/) -name ptxas -print -quit) $VIRTUAL_ENV/bin/ptxas ``` Verify the GPU setup: ``` python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` ### 1\. System requirements - macOS 10.12.6 (Sierra) or higher (64-bit) Note:* While TensorFlow supports Apple Silicon (M1), packages that include custom C++ extensions for TensorFlow also need to be compiled for Apple M1. Some packages, like [tensorflow\_decision\_forests](https://www.tensorflow.org/decision_forests) publish M1-compatible versions, but many packages don't. To use those libraries, you will have to use TensorFlow with x86 emulation and Rosetta. Currently there is no official GPU support for running TensorFlow on MacOS. The following instructions are for running on CPU. ### 2\. Check Python version Check if your Python environment is already configured: Note: Requires Python 3.9–3.11, and pip \>= 20.3 for MacOS. ``` python3 --version python3 -m pip --version ``` ### 3\. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. ``` pip install --upgrade pip ``` Then, install TensorFlow with pip. ``` pip install tensorflow ``` ### 4\. Verify the installation ``` python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` If a tensor is returned, you've installed TensorFlow successfully. Caution: TensorFlow `2.10` was the last TensorFlow release that supported GPU on native-Windows. Starting with TensorFlow `2.11`, you will need to install [TensorFlow in WSL2](https://tensorflow.org/install/pip#windows-%5Bwsl2%5D), or install `tensorflow-cpu` and, optionally, try the [TensorFlow-DirectML-Plugin](https://github.com/microsoft/tensorflow-directml-plugin#tensorflow-directml-plugin-) ## 1\. System requirements - Windows 7 or higher (64-bit) Note: Starting with TensorFlow `2.10`, Windows CPU-builds for x86/x64 processors are built, maintained, tested and released by a third party: [Intel](https://www.intel.com/). Installing the windows-native [`tensorflow`](https://pypi.org/project/tensorflow/) or [`tensorflow-cpu`](https://pypi.org/project/tensorflow-cpu/) package installs Intel's [`tensorflow-intel`](https://pypi.org/project/tensorflow-intel/) package. These packages are provided as-is. Tensorflow will use reasonable efforts to maintain the availability and integrity of this pip package. There may be delays if the third party fails to release the pip package. See [this blog post](https://blog.tensorflow.org/2022/09/announcing-tensorflow-official-build-collaborators.html) for more information about this collaboration. ### 2\. Install Microsoft Visual C++ Redistributable Install the Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017, and 2019. Starting with the TensorFlow 2.1.0 version, the `msvcp140_1.dll` file is required from this package (which may not be provided from older redistributable packages). The redistributable comes with Visual Studio 2019 but can be installed separately: 1. Go to the [Microsoft Visual C++ downloads](https://support.microsoft.com/help/2977003/the-latest-supported-visual-c-downloads). 2. Scroll down the page to the Visual Studio 2015, 2017 and 2019 section. 3. Download and install the Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019 for your platform. Make sure [long paths are enabled](https://superuser.com/questions/1119883/windows-10-enable-ntfs-long-paths-policy-option-missing) on Windows. ### 3\. Install Miniconda [Miniconda](https://docs.conda.io/en/latest/miniconda.html) is the recommended approach for installing TensorFlow with GPU support. It creates a separate environment to avoid changing any installed software in your system. This is also the easiest way to install the required software especially for the GPU setup. Download the [Miniconda Windows Installer](https://repo.anaconda.com/miniconda/Miniconda3-latest-Windows-x86_64.exe). Double-click the downloaded file and follow the instructions on the screen. ### 4\. Create a conda environment Create a new conda environment named [`tf`](https://www.tensorflow.org/api_docs/python/tf) with the following command. ``` conda create --name tf python=3.9 ``` You can deactivate and activate it with the following commands. ``` conda deactivate conda activate tf ``` Make sure it is activated for the rest of the installation. ### 5\. GPU setup You can skip this section if you only run TensorFlow on CPU. First install [NVIDIA GPU driver](https://www.nvidia.com/Download/index.aspx) if you have not. Then install the CUDA, cuDNN with conda. ``` conda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0 ``` ### 6\. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. ``` pip install --upgrade pip ``` Then, install TensorFlow with pip. Note: Do not install TensorFlow with conda. It may not have the latest stable version. pip is recommended since TensorFlow is only officially released to PyPI. ``` # Anything above 2.10 is not supported on the GPU on Windows Native pip install "tensorflow<2.11" ``` ### 7\. Verify the installation Verify the CPU setup: ``` python -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` If a tensor is returned, you've installed TensorFlow successfully. Verify the GPU setup: ``` python -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` If a list of GPU devices is returned, you've installed TensorFlow successfully. ### 1\. System requirements - Windows 10 19044 or higher (64-bit). This corresponds to Windows 10 version 21H2, the November 2021 update. See the following documents to: - [Download the latest Windows 10 update](https://www.microsoft.com/software-download/windows10). - [Install WSL2](https://docs.microsoft.com/windows/wsl/install) - [Setup NVIDIA® GPU support in WSL2](https://docs.nvidia.com/cuda/wsl-user-guide/index.html) ### 2\. GPU setup You can skip this section if you only run TensorFlow on the CPU. Install the [NVIDIA GPU driver](https://www.nvidia.com/Download/index.aspx) if you have not. You can use the following command to verify it is installed. ``` nvidia-smi ``` ### 3\. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. ``` pip install --upgrade pip ``` Then, install TensorFlow with pip. ``` # For GPU users pip install tensorflow[and-cuda] # For CPU users pip install tensorflow ``` ### 4\. Verify the installation Verify the CPU setup: ``` python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` If a tensor is returned, you've installed TensorFlow successfully. Verify the GPU setup: ``` python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` If a list of GPU devices is returned, you've installed TensorFlow successfully. ## Package location A few installation mechanisms require the URL of the TensorFlow Python package. The value you specify depends on your Python version. ## Python version support Warning: As of TensorFlow 2.21, Python 3.9 is no longer supported. Please use a supported Python version (e.g., 3.10-3.13). \| Version \| URL \| \|---\|---\| \| Linux x86 \| \| \| Python 3.10 GPU support \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl) \| \| Python 3.10 CPU-only \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow\_cpu-2.21.0-cp310-cp310-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow_cpu-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl) \| \| Python 3.11 GPU support \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl) \| \| Python 3.11 CPU-only \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow\_cpu-2.21.0-cp311-cp311-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow_cpu-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl) \| \| Python 3.12 GPU support \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_x86_64.whl) \| \| Python 3.12 CPU-only \| <https://storage.googleapis.com/tensorflow/versions/2.20.0/tensorflow_cpu-2.20.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl> \| \| Python 3.13 GPU support \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl) \| \| Python 3.13 CPU-only \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow\_cpu-2.21.0-cp313-cp313-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow_cpu-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl) \| \| Linux Arm64 (CPU-only) \| \| \| Python 3.10 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-manylinux\_2\_27\_aarch64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_aarch64.whl) \| \| Python 3.11 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-manylinux\_2\_27\_aarch64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_aarch64.whl) \| \| Python 3.12 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-manylinux\_2\_27\_aarch64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_aarch64.whl) \| \| Python 3.13 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-manylinux\_2\_27\_aarch64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_aarch64.whl) \| \| macOS x86 (CPU-only) \| \| \| Caution: TensorFlow 2.16 was the last TensorFlow release that supported macOS x86 \| \| \| Python 3.10 \| <https://storage.googleapis.com/tensorflow/versions/2.16.2/tensorflow-2.16.2-cp310-cp310-macosx_10_15_x86_64.whl> \| \| Python 3.11 \| <https://storage.googleapis.com/tensorflow/versions/2.16.2/tensorflow-2.16.2-cp311-cp311-macosx_10_15_x86_64.whl> \| \| Python 3.12 \| <https://storage.googleapis.com/tensorflow/versions/2.16.2/tensorflow-2.16.2-cp312-cp312-macosx_10_15_x86_64.whl> \| \| macOS Arm64 (CPU-only) \| \| \| Python 3.10 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-macosx\_12\_0\_arm64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-macosx_12_0_arm64.whl) \| \| Python 3.11 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-macosx\_12\_0\_arm64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-macosx_12_0_arm64.whl) \| \| Python 3.12 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-macosx\_12\_0\_arm64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-macosx_12_0_arm64.whl) \| \| Python 3.13 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-macosx\_12\_0\_arm64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-macosx_12_0_arm64.whl) \| \| Windows (CPU-only) \| \| \| Python 3.10 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-win\_amd64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-win_amd64.whl) \| \| Python 3.11 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-win\_amd64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-win_amd64.whl) \| \| Python 3.12 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-win\_amd64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-win_amd64.whl) \| \| Python 3.13 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-win\_amd64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-win_amd64.whl) \| Was this helpful? Except as otherwise noted, the content of this page is licensed under the [Creative Commons Attribution 4.0 License](https://creativecommons.org/licenses/by/4.0/), and code samples are licensed under the [Apache 2.0 License](https://www.apache.org/licenses/LICENSE-2.0). For details, see the [Google Developers Site Policies](https://developers.google.com/site-policies). Java is a registered trademark of Oracle and/or its affiliates. Last updated 2026-03-12 UTC. - ### Stay connected - [Blog](https://blog.tensorflow.org/) - [Forum](https://discuss.tensorflow.org/) - [GitHub](https://github.com/tensorflow/) - [Twitter](https://twitter.com/tensorflow) - [YouTube](https://youtube.com/tensorflow) - ### Support - [Issue tracker](https://github.com/tensorflow/tensorflow/issues) - [Release notes](https://github.com/tensorflow/tensorflow/blob/master/RELEASE.md) - [Stack Overflow](https://stackoverflow.com/questions/tagged/tensorflow) - [Brand guidelines](https://www.tensorflow.org/extras/tensorflow_brand_guidelines.pdf) - [Cite TensorFlow](https://www.tensorflow.org/about/bib) - [Terms](https://policies.google.com/terms) - [Privacy](https://policies.google.com/privacy) - [Manage cookies](https://www.tensorflow.org/install/pip) - Sign up for the TensorFlow newsletter [Subscribe](https://www.tensorflow.org/subscribe) - [English](https://www.tensorflow.org/install/pip) - [Español](https://www.tensorflow.org/install/pip?hl=es) - [Español – América Latina](https://www.tensorflow.org/install/pip?hl=es-419) - [Français](https://www.tensorflow.org/install/pip?hl=fr) - [Indonesia](https://www.tensorflow.org/install/pip?hl=id) - [Italiano](https://www.tensorflow.org/install/pip?hl=it) - [Polski](https://www.tensorflow.org/install/pip?hl=pl) - [Português](https://www.tensorflow.org/install/pip?hl=pt) - [Português – Brasil](https://www.tensorflow.org/install/pip?hl=pt-br) - [Tiếng Việt](https://www.tensorflow.org/install/pip?hl=vi) - [Türkçe](https://www.tensorflow.org/install/pip?hl=tr) - [Русский](https://www.tensorflow.org/install/pip?hl=ru) - [עברית](https://www.tensorflow.org/install/pip?hl=he) - [العربيّة](https://www.tensorflow.org/install/pip?hl=ar) - [فارسی](https://www.tensorflow.org/install/pip?hl=fa) - [हिंदी](https://www.tensorflow.org/install/pip?hl=hi) - [বাংলা](https://www.tensorflow.org/install/pip?hl=bn) - [ภาษาไทย](https://www.tensorflow.org/install/pip?hl=th) - [中文 – 简体](https://www.tensorflow.org/install/pip?hl=zh-cn) - [中文 – 繁體](https://www.tensorflow.org/install/pip?hl=zh-tw) - [日本語](https://www.tensorflow.org/install/pip?hl=ja) - [한국어](https://www.tensorflow.org/install/pip?hl=ko)
Readable Markdown	## Install TensorFlow with pipStay organized with collections Save and categorize content based on your preferences. - On this page - [Hardware requirements](https://www.tensorflow.org/install/pip#hardware_requirements) - [System requirements](https://www.tensorflow.org/install/pip#system_requirements) - [Software requirements](https://www.tensorflow.org/install/pip#software_requirements) - [Step-by-step instructions](https://www.tensorflow.org/install/pip#step-by-step_instructions) - [Package location](https://www.tensorflow.org/install/pip#package_location) - [Python version support](https://www.tensorflow.org/install/pip#python_version_support) This guide is for the latest stable version of TensorFlow. For the preview build (nightly), use the pip package named `tf-nightly`. Refer to [these tables](https://www.tensorflow.org/install/source#tested_build_configurations) for older TensorFlow version requirements. For the CPU-only build, use the pip package named `tensorflow-cpu`. Here are the quick versions of the install commands. Scroll down for the step-by-step instructions. ``` python3 -m pip install 'tensorflow[and-cuda]' # Verify the installation: python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` ``` # There is currently no official GPU support for MacOS. python3 -m pip install tensorflow # Verify the installation: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` ``` conda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0 # Anything above 2.10 is not supported on the GPU on Windows Native python -m pip install "tensorflow<2.11" # Verify the installation: python -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` ``` python3 -m pip install tensorflow[and-cuda] # Verify the installation: python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` ``` python3 -m pip install tensorflow # Verify the installation: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` ``` python3 -m pip install tf-nightly # Verify the installation: python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` ## Hardware requirements The following GPU-enabled devices are supported: - NVIDIA® GPU card with CUDA® architectures 3.5, 5.0, 6.0, 7.0, 7.5, 8.0 and higher. See the list of [CUDA®-enabled GPU cards](https://developer.nvidia.com/cuda-gpus). - For GPUs with unsupported CUDA® architectures, or to avoid JIT compilation from PTX, or to use different versions of the NVIDIA® libraries, see the [Linux build from source](https://www.tensorflow.org/install/source) guide. - Packages do not contain PTX code except for the latest supported CUDA® architecture; therefore, TensorFlow fails to load on older GPUs when `CUDA_FORCE_PTX_JIT=1` is set. (See [Application Compatibility](https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#application-compatibility) for details.) ## System requirements - Ubuntu 16.04 or higher (64-bit) - macOS 12.0 (Monterey) or higher (64-bit) (no GPU support) - Windows Native - Windows 7 or higher (64-bit) (no GPU support after TF 2.10) - Windows WSL2 - Windows 10 19044 or higher (64-bit) ## Software requirements - Python 3.9–3.12 - pip version 19.0 or higher for Linux (requires `manylinux2014` support) and Windows. pip version 20.3 or higher for macOS. - Windows Native Requires [Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019](https://learn.microsoft.com/en-us/cpp/windows/latest-supported-vc-redist) The following NVIDIA® software are only required for GPU support. - [NVIDIA® GPU drivers](https://www.nvidia.com/drivers) - \>= 525.60.13 for Linux - \>= 528.33 for WSL on Windows - [CUDA® Toolkit 12.3](https://developer.nvidia.com/cuda-toolkit-archive). - [cuDNN SDK 8.9.7](https://developer.nvidia.com/cudnn). - (Optional) [TensorRT](https://docs.nvidia.com/deeplearning/tensorrt/archives/index.html#trt_7) to improve latency and throughput for inference. ## Step-by-step instructions ### 1\. System requirements - Ubuntu 16.04 or higher (64-bit) TensorFlow only officially supports Ubuntu. However, the following instructions may also work for other Linux distros. ### 2\. GPU setup You can skip this section if you only run TensorFlow on the CPU. Install the [NVIDIA GPU driver](https://www.nvidia.com/Download/index.aspx) if you have not. You can use the following command to verify it is installed. ``` nvidia-smi ``` ### 3\. Create a virtual environment with [venv](https://docs.python.org/3/library/venv.html) The venv module is part of Python’s standard library and is the officially recommended way to create virtual environments. Navigate to your desired virtual environments directory and create a new venv environment named [`tf`](https://www.tensorflow.org/api_docs/python/tf) with the following command. ``` python3 -m venv tf ``` You can activate it with the following command. ``` source tf/bin/activate ``` Make sure that the virtual environment is activated for the rest of the installation. ### 4\. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. ``` pip install --upgrade pip ``` Then, install TensorFlow with pip. ``` # For GPU users pip install tensorflow[and-cuda] # For CPU users pip install tensorflow ``` ### 6\. Verify the installation Verify the CPU setup: ``` python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` If a tensor is returned, you've installed TensorFlow successfully. Verify the GPU setup: ``` python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` If a list of GPU devices is returned, you've installed TensorFlow successfully. If not continue to the next step. ### 6\. \[GPU only\] Virtual environment configuration If the GPU test in the last section was unsuccessful, the most likely cause is that components aren't being detected, and/or conflict with the existing system CUDA installation. So you need to add some symbolic links to fix this. - Create symbolic links to NVIDIA shared libraries: ``` pushd $(dirname $(python -c 'print(__import__("tensorflow").__file__)')) ln -svf ../nvidia//lib/.so* . popd ``` - Create a symbolic link to ptxas: ``` ln -sf $(find $(dirname $(dirname $(python -c "import nvidia.cuda_nvcc; print(nvidia.cuda_nvcc.__file__)"))//bin/) -name ptxas -print -quit) $VIRTUAL_ENV/bin/ptxas ``` Verify the GPU setup: ``` python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` ### 1\. System requirements - macOS 10.12.6 (Sierra) or higher (64-bit) Currently there is no official GPU support for running TensorFlow on MacOS. The following instructions are for running on CPU. ### 2\. Check Python version Check if your Python environment is already configured: ``` python3 --version python3 -m pip --version ``` ### 3\. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. ``` pip install --upgrade pip ``` Then, install TensorFlow with pip. ``` pip install tensorflow ``` ### 4\. Verify the installation ``` python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` If a tensor is returned, you've installed TensorFlow successfully. ## 1\. System requirements - Windows 7 or higher (64-bit) ### 2\. Install Microsoft Visual C++ Redistributable Install the Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017, and 2019. Starting with the TensorFlow 2.1.0 version, the `msvcp140_1.dll` file is required from this package (which may not be provided from older redistributable packages). The redistributable comes with Visual Studio 2019* but can be installed separately: 1. Go to the [Microsoft Visual C++ downloads](https://support.microsoft.com/help/2977003/the-latest-supported-visual-c-downloads). 2. Scroll down the page to the Visual Studio 2015, 2017 and 2019 section. 3. Download and install the Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019 for your platform. Make sure [long paths are enabled](https://superuser.com/questions/1119883/windows-10-enable-ntfs-long-paths-policy-option-missing) on Windows. ### 3\. Install Miniconda [Miniconda](https://docs.conda.io/en/latest/miniconda.html) is the recommended approach for installing TensorFlow with GPU support. It creates a separate environment to avoid changing any installed software in your system. This is also the easiest way to install the required software especially for the GPU setup. Download the [Miniconda Windows Installer](https://repo.anaconda.com/miniconda/Miniconda3-latest-Windows-x86_64.exe). Double-click the downloaded file and follow the instructions on the screen. ### 4\. Create a conda environment Create a new conda environment named [`tf`](https://www.tensorflow.org/api_docs/python/tf) with the following command. ``` conda create --name tf python=3.9 ``` You can deactivate and activate it with the following commands. ``` conda deactivate conda activate tf ``` Make sure it is activated for the rest of the installation. ### 5\. GPU setup You can skip this section if you only run TensorFlow on CPU. First install [NVIDIA GPU driver](https://www.nvidia.com/Download/index.aspx) if you have not. Then install the CUDA, cuDNN with conda. ``` conda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0 ``` ### 6\. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. ``` pip install --upgrade pip ``` Then, install TensorFlow with pip. ``` # Anything above 2.10 is not supported on the GPU on Windows Native pip install "tensorflow<2.11" ``` ### 7\. Verify the installation Verify the CPU setup: ``` python -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` If a tensor is returned, you've installed TensorFlow successfully. Verify the GPU setup: ``` python -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` If a list of GPU devices is returned, you've installed TensorFlow successfully. ### 1\. System requirements - Windows 10 19044 or higher (64-bit). This corresponds to Windows 10 version 21H2, the November 2021 update. See the following documents to: - [Download the latest Windows 10 update](https://www.microsoft.com/software-download/windows10). - [Install WSL2](https://docs.microsoft.com/windows/wsl/install) - [Setup NVIDIA® GPU support in WSL2](https://docs.nvidia.com/cuda/wsl-user-guide/index.html) ### 2\. GPU setup You can skip this section if you only run TensorFlow on the CPU. Install the [NVIDIA GPU driver](https://www.nvidia.com/Download/index.aspx) if you have not. You can use the following command to verify it is installed. ``` nvidia-smi ``` ### 3\. Install TensorFlow TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version. ``` pip install --upgrade pip ``` Then, install TensorFlow with pip. ``` # For GPU users pip install tensorflow[and-cuda] # For CPU users pip install tensorflow ``` ### 4\. Verify the installation Verify the CPU setup: ``` python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))" ``` If a tensor is returned, you've installed TensorFlow successfully. Verify the GPU setup: ``` python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))" ``` If a list of GPU devices is returned, you've installed TensorFlow successfully. ## Package location A few installation mechanisms require the URL of the TensorFlow Python package. The value you specify depends on your Python version. ## Python version support \| Version \| URL \| \|---\|---\| \| Linux x86 \| \| \| Python 3.10 GPU support \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl) \| \| Python 3.10 CPU-only \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow\_cpu-2.21.0-cp310-cp310-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow_cpu-2.21.0-cp310-cp310-manylinux_2_27_x86_64.whl) \| \| Python 3.11 GPU support \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl) \| \| Python 3.11 CPU-only \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow\_cpu-2.21.0-cp311-cp311-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow_cpu-2.21.0-cp311-cp311-manylinux_2_27_x86_64.whl) \| \| Python 3.12 GPU support \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_x86_64.whl) \| \| Python 3.12 CPU-only \| <https://storage.googleapis.com/tensorflow/versions/2.20.0/tensorflow_cpu-2.20.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl> \| \| Python 3.13 GPU support \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl) \| \| Python 3.13 CPU-only \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow\_cpu-2.21.0-cp313-cp313-manylinux\_2\_27\_x86\_64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow_cpu-2.21.0-cp313-cp313-manylinux_2_27_x86_64.whl) \| \| Linux Arm64 (CPU-only) \| \| \| Python 3.10 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-manylinux\_2\_27\_aarch64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-manylinux_2_27_aarch64.whl) \| \| Python 3.11 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-manylinux\_2\_27\_aarch64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-manylinux_2_27_aarch64.whl) \| \| Python 3.12 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-manylinux\_2\_27\_aarch64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-manylinux_2_27_aarch64.whl) \| \| Python 3.13 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-manylinux\_2\_27\_aarch64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-manylinux_2_27_aarch64.whl) \| \| macOS x86 (CPU-only) \| \| \| Caution: TensorFlow 2.16 was the last TensorFlow release that supported macOS x86 \| \| \| Python 3.10 \| <https://storage.googleapis.com/tensorflow/versions/2.16.2/tensorflow-2.16.2-cp310-cp310-macosx_10_15_x86_64.whl> \| \| Python 3.11 \| <https://storage.googleapis.com/tensorflow/versions/2.16.2/tensorflow-2.16.2-cp311-cp311-macosx_10_15_x86_64.whl> \| \| Python 3.12 \| <https://storage.googleapis.com/tensorflow/versions/2.16.2/tensorflow-2.16.2-cp312-cp312-macosx_10_15_x86_64.whl> \| \| macOS Arm64 (CPU-only) \| \| \| Python 3.10 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-macosx\_12\_0\_arm64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-macosx_12_0_arm64.whl) \| \| Python 3.11 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-macosx\_12\_0\_arm64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-macosx_12_0_arm64.whl) \| \| Python 3.12 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-macosx\_12\_0\_arm64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-macosx_12_0_arm64.whl) \| \| Python 3.13 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-macosx\_12\_0\_arm64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-macosx_12_0_arm64.whl) \| \| Windows (CPU-only) \| \| \| Python 3.10 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-win\_amd64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp310-cp310-win_amd64.whl) \| \| Python 3.11 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-win\_amd64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp311-cp311-win_amd64.whl) \| \| Python 3.12 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-win\_amd64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp312-cp312-win_amd64.whl) \| \| Python 3.13 \| [https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-win\_amd64.whl](https://storage.googleapis.com/tensorflow/versions/%222.21.0%22/tensorflow-2.21.0-cp313-cp313-win_amd64.whl) \| Except as otherwise noted, the content of this page is licensed under the [Creative Commons Attribution 4.0 License](https://creativecommons.org/licenses/by/4.0/), and code samples are licensed under the [Apache 2.0 License](https://www.apache.org/licenses/LICENSE-2.0). For details, see the [Google Developers Site Policies](https://developers.google.com/site-policies). Java is a registered trademark of Oracle and/or its affiliates. Last updated 2026-03-12 UTC.
Shard	168 (laksa)
Root Hash	12537842311192732768
Unparsed URL	org,tensorflow!www,/install/pip s443