Could not find version that satisfies the requirement tensorflow

# Could not find version that satisfies the requirement tensorflow

The error "Could not find a version that satisfies the requirement tensorflow" occurs for multiple reasons:

• Using a 32-bit version of Python as TensorFlow only supports 64-bit Python
• Having an outdated version of pip or using pip for Python 2 instead of pip3.
• Using a Python version that is not supported by TensorFlow.

Copied!
Collecting tensorflow
Could not find a version that satisfies the requirement tensorflow (from versions: )
No matching distribution found for tensorflow.

# Upgrade your version of pip

The first thing to do is to upgrade your pip version.

Here are the commands for upgrading pip on all operating systems.

Which command works depends on your operating system and your version of Python.

Copied!
# 👇️ If you have pip already installed
pip install --upgrade pip

# 👇️ If your pip is aliased as pip3 (Python 3)
pip3 install --upgrade pip

# 👇️ If you don't have pip in your PATH environment variable
python -m pip install --upgrade pip

# 👇️ If you don't have pip in your PATH environment variable
python3 -m pip install --upgrade pip

# 👇️ If you have easy_install
easy_install --upgrade pip

# 👇️ If you get a permissions error
sudo easy_install --upgrade pip

# 👇️ If you get a permissions error when upgrading pip
pip install --upgrade pip --user

# 👇️ Upgrade pip scoped to the current user (if you get a permissions error)
python -m pip install --user --upgrade pip
python3 -m pip install --user --upgrade pip

# 👇️ Installing directly from get-pip.py (MacOS and Linux)
curl https://bootstrap.pypa.io/get-pip.py | python

# 👇️ If you get permissions issues
curl https://bootstrap.pypa.io/get-pip.py | sudo python

# 👇️ Alternative for Ubuntu/Debian
sudo apt-get update && apt-get upgrade python-pip

# 👇️ Alternative for Red Hat / CentOS / Fedora
sudo yum install epel-release
sudo yum install python-pip
sudo yum update python-pip

After you upgrade pip, try installing tensorflow.

Copied!
# 👇️ In a virtual environment or using Python 2
pip install tensorflow

# 👇️ For python 3 (could also be pip3.10 depending on your version)
pip3 install tensorflow

# 👇️ If you get a permissions error
sudo pip3 install tensorflow

# 👇️ If you don't have pip in your PATH environment variable
python -m pip install tensorflow

# 👇️ For python 3 (could also be pip3.10 depending on your version)
python3 -m pip install tensorflow

# 👇️ For Anaconda
conda install -c conda-forge tensorflow

If the error persists, try installing tensorflow from the official URLs of the TensorFlow package.

The value you specify depends on your Python version. You can view all of the URLS in this table in the docs.

Here is an example that uses the Python 3.10 installation URLs.

You can replace python3 with python if you have it aliased.

Copied!
# 👇️ For Windows
python3 -m pip install --upgrade https://storage.googleapis.com/tensorflow/windows/cpu/tensorflow_cpu-2.10.0-cp310-cp310-win_amd64.whl

# 👇️ For macOS
python3 -m pip install --upgrade https://storage.googleapis.com/tensorflow/mac/cpu/tensorflow-2.10.0-cp310-cp310-macosx_10_14_x86_64.whl

# 👇️ For Linux
python3 -m pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow_cpu-2.10.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

The official installation URLs for all of the supported by TensorFlow Python versions are available in this table of the docs.

# TensorFlow has specific requirements

The TensorFlow module has a couple of requirements.

The module only supports 64-bit Python.

Run the following command in your shell to check if your Python is 64-bit.

Copied!
python -c "import sys; print(sys.maxsize > 2**32)"

If the output of the command is True, then your Python is 64-bit.

The TensorFlow module also has a specific range of Python versions it supports.

Here are TensorFlow's requirements at the time of writing:

1. Python version 3.7-3.10 (Must be 64-bit)
2. pip version 19.0 or higher for Linux and Windows
3. pip version 20.3 or higher for macOS

Here is the list of requirements in the official documentation.

You can check your Python version by running the following command.

Copied!
python --version

You can download a supported Python version from the official downloads page.

Make sure the Python package is 64-bit and is in the supported by TensorFlow version range.

Make sure to tick the following options if you get prompted:

• Install launcher for all users (recommended)
• Add Python to PATH (this adds Python to your PATH environment variable)

If you use Anaconda, you can install a specific Python version using the following command:

Copied!
conda install python=3.10.6

There is a very good chance that the latest Python version is supported by TensorFlow, but check the list of requirements to be sure.

If necessary, you can download a specific Python version that is supported by the package if it doesn't support the latest Python version.

Different versions are available in the "Looking for a specific release" table.

After you download a supported by TensorFlow version, try running the pip install tensorflow command.

Copied!
# 👇️ Requires the latest pip
pip install --upgrade pip

# 👇️ In a virtual environment or using Python 2
pip install tensorflow

# 👇️ For python 3 (could also be pip3.10 depending on your version)
pip3 install tensorflow

# 👇️ If you get a permissions error
sudo pip3 install tensorflow

# 👇️ If you don't have pip in your PATH environment variable
python -m pip install tensorflow

# 👇️ For python 3 (could also be pip3.10 depending on your version)
python3 -m pip install tensorflow

# 👇️ For Anaconda
conda install -c conda-forge tensorflow

If that didn't work, try running the command with python -m.

# Install tensorflow with "python -m pip install tensorflow"

Try running the pip install command as python -m pip install tensorflow.

This is necessary if you don't have pip in your PATH environment variable.

Copied!
# 👇️ If you don't have pip in your PATH environment variable
python -m pip install tensorflow

# 👇️ For Python 3 outside of a virtual environment
python3 -m pip install tensorflow

You can also try to install the package with pip3 instead of pip.

Copied!
pip3 install tensorflow

If that didn't solve the error, try installing the package scoped to the user.

# Install the package with the --user option

The error is often caused due to not having the necessary permissions to install the package for all users on the machine.

To solve the error, install the package scoped to the specific user with the --user option.

Copied!
pip install tensorflow --user

pip3 install tensorflow --user

python3 -m pip install tensorflow --user

The --user option installs the package in the user's home directory.

If you get a permissions error, try running the command with the --user flag or with sudo.

Copied!
sudo pip install tensorflow

sudo pip3 install tensorflow

sudo python3 -m pip install tensorflow

If that didn't solve the error, try upgrading the module.

# Upgrade the package with the --upgrade option

The error is also caused if you try to install an older version of the package that doesn't support your version of Python.

If that is the case, try upgrading the version of the package.

Copied!
pip install tensorflow --upgrade

pip3 install tensorflow --upgrade

python3 -m pip install tensorflow --upgrade

You can also try to install a specific version of the module.

Copied!
python3 -m pip install --pre --upgrade tensorflow==XX.XX.XX

You have to specify a version number after the equal signs.

Copied!
pip install tensorflow==

pip3 install tensorflow==

Here is a screenshot of issuing the command in my terminal.

The output contains a tuple of all of the versions of the package from the oldest to the most recent.

For example, if I wanted to install the version 2.10.0, I would run the following command.

Copied!
python3 -m pip install --pre --upgrade tensorflow==2.10.0

# If installing from a requirements.txt file, make sure it contains the dependencies

If you are installing from a requirements.txt file with pip install -r requirements.txt, you have to make sure the file contains all of the packages and their dependencies.

This can be done by running pip install <package-name> and generating a new requirements.txt file that contains all of the dependencies.

Copied!
pip freeze > requirements.txt

pip3 freeze > requirements.txt

The error commonly occurs if your requirements.txt file contains only some of the dependencies of the packages.

You can also add the dependencies of the packages manually to your requirements.txt file based on the output from the error message.

# Try running the pip install tensorflow command with the --pre option

The --pre option makes it so pip includes pre-release and development versions of the package. By default pip only finds stable versions.

Copied!
pip install tensorflow --pre

pip3 install tensorflow --pre

python -m pip install tensorflow --pre

python3 -m pip install tensorflow --pre

# Try installing from the official installation URLs

If the error persists, try installing tensorflow from the official installation URLs.

In some cases installing from the URLs is your only option.

The value you specify depends on your Python version. You can view all of the URLS in this table in the docs.

Here is an example that uses the Python 3.10 installation URLs.

You can replace python3 with python if you have it aliased.

Copied!
# 👇️ For Windows
python3 -m pip install --upgrade https://storage.googleapis.com/tensorflow/windows/cpu/tensorflow_cpu-2.10.0-cp310-cp310-win_amd64.whl

# 👇️ For macOS
python3 -m pip install --upgrade https://storage.googleapis.com/tensorflow/mac/cpu/tensorflow-2.10.0-cp310-cp310-macosx_10_14_x86_64.whl

# 👇️ For Linux
python3 -m pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow_cpu-2.10.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

The official installation URLs for all of the supported by TensorFlow Python versions are available in this table of the docs.

# Try installing tensorflow in a virtual environment

Another thing that might help is to create a virtual environment if you don't already have one.

Copied!
# 👇️ Use the correct version of Python when creating VENV
python3 -m venv venv

# 👇️ Activate on Unix or MacOS
source venv/bin/activate

# 👇️ Activate on Windows (cmd.exe)
venv\Scripts\activate.bat

# 👇️ Activate on Windows (PowerShell)
venv\Scripts\Activate.ps1

# 👇️ Requires the latest pip
pip install --upgrade pip

# 👇️ Install tensorflow in your virtual environment
pip install tensorflow

Make sure to use the correct activation command depending on your operating system.

Your virtual environment will use the version of Python that was used to create it.

# Try running pip install in verbose mode

If none of the suggestions helped, try running the pip install command in verbose mode.

Copied!
pip install tensorflow -vvv

pip3 install tensorflow -vvv

python -m pip install tensorflow -vvv

The -v option stands for verbose mode and can be used up to 3 times.

When the pip install command is run in verbose mode, the command shows more output and how the error occurred.

# Conclusion

To solve the error "Could not find a version that satisfies the requirement tensorflow", make sure:

1. To use a 64-bit version of Python.
2. To have an up-to-date version of pip installed.
3. To use a Python version that is supported by TensorFlow.
4. You have the necessary permissions to install the package.

# Additional Resources

You can learn more about the related topics by checking out the following tutorials:

• RuntimeError: tf.placeholder() is not compatible with eager execution
• How to disable/suppress Tensorflow warnings in Python
• Disable specific Pylint warnings or all warnings in a File
• ModuleNotFoundError: No module named 'tensorflow' in Python
• Python: Can't call numpy() on Tensor that requires grad