API keys - Anaconda

API keys authenticate users and automated systems to access platform resources. You can include multiple scopes to define access permissions in API keys that you create, but you cannot grant access to resources that you do not have access to already.

Creating an API Key

Sign in to Anaconda.
Open the user dropdown menu and select Account Settings.
Select the API Keys tab.
Select Create API Key.
Provide a unique name for your API key.
Choose permission levels for the scopes your key should include. You can select one or both of the following:
- User resources: Provides access to the platform’s user-level APIs. Include this scope if you need to interact programmatically with projects, data, or other assets associated with your account.
  This scope mirrors your platform permissions (except for any administrative controls).
- Platform channels: Authenticates your local conda client or automated build systems so they can use channels you have access to within Anaconda Platform. Include this scope if you need to install packages from channels hosted on your instance.
Set an expiration date from the dropdown. API keys have a maximum lifespan of one year.
If you choose Custom, a calendar field appears. Use it to specify a date.
Select Create API Key. Save your newly generated API key in a secure location.
Alternatively, you can generate an API key by logging in at the command line using anaconda-auth login.

Using API Keys

If your conda client runs as part of an automated process, such as a CI/CD pipeline, or if you are using conda within Docker builds on your desktop, configure them to use your local API key (generated when you logged in using anaconda-auth) to authenticate conda using one of the following methods:

Using an environment variable
Using the keyring
Docker

Open Anaconda Prompt or Powershell (Terminal on macOS/Linux).
Set your API key as an environment variable:
for /f "usebackq delims=" %i in (`anaconda auth api-key`) do set ANACONDA_AUTH_API_KEY=%i

Verify preferred_token_storage is set to "anaconda-keyring". See anaconda-auth configuration parameters.
Open Anaconda Prompt (Terminal on macOS/Linux).
Copy the ~/.anaconda/keyring file to your automated workflow system:
```
scp ~/.anaconda/keyring user@<TARGET_SYSTEM>:~/.anaconda/keyring
```

If you use conda in Docker builds, you can securely provide your Anaconda API key during build time using Docker secrets. This allows authenticated commands to run without exposing your API key in an image layer.

Open Anaconda Prompt or Powershell (Terminal on macOS/Linux).
Set your API key as an environment variable:
for /f "usebackq delims=" %i in (`anaconda auth api-key`) do set ANACONDA_AUTH_API_KEY=%i

Each Dockerfile must also contain the anaconda-registration package:

RUN conda install --name base anaconda-registration --channel https://repo.anaconda.cloud/repo/anaconda-tools --override-channels

For all other conda or anaconda commands that require authentication, provide the following:
```
RUN --mount=type=secret,id=ANACONDA_AUTH_API_KEY 
```
Example command requiring auth
```
RUN --mount=type=secret,id=ANACONDA_AUTH_API_KEY conda create -p /opt/env -y "python=3.13" "anaconda-client"
```
When building your image, pass the key as a Docker secret:
```
docker build --secret id=ANACONDA_AUTH_API_KEY -t <IMAGE_NAME> .
```
Replace <IMAGE_NAME> with the name for your Docker image.

For API integrations, include the key in the authorization header:

Authorization: Bearer (anaconda auth api-key)

This lets automated workflows—such as data synchronization jobs or deployment pipelines—interact securely with the platform using the same permissions you hold.

Anaconda Platform (Cloud)

​Creating an API Key

​Using API Keys

Creating an API Key

Using API Keys