Skip to main content
Choose how customers access your model. You can allowlist Model API access for selected customer workspaces, make a Model API broadly available after validation and assisted setup, or offer Dedicated availability. Baseten configures Dedicated availability during assisted setup, without an activation URL. For Model APIs, send the customer-facing activation steps to an administrator of the customer’s workspace. That administrator accepts the required terms in their own Baseten workspace.

​
Choose who can access your model

Allowlisted Model API access limits the activation handoff to the customer workspaces you select. For broad Model API availability, complete validation and assisted setup with Baseten, then send the activation handoff to each customer workspace. For Dedicated availability, Baseten configures customer availability during assisted setup.

​
Give customers access to a Model API

Each Model API has an activation URL with this format:
Your Baseten contact provides the activation URL. The <SharedEndpointId> value identifies the Model API. The URL activates the model only in the workspace whose admin accepts the terms. To share a customer activation link:
  1. Copy the activation URL provided by Baseten.
  2. Send the URL unchanged to an admin of the customer’s Baseten workspace.
Only a workspace admin can accept the agreements for the workspace. Give that admin the following steps. To activate a Model API:
  1. Sign in to your workspace at app.baseten.co.
  2. Open the activation URL shared by the model provider.
  3. On Add <model name>, review the model details.
  4. Under Information shared with <provider>, choose I agree to share this information with <provider>.
  5. Under Terms and agreements, choose Review agreement.
  6. If the provider supplies a terms URL, choose Review terms.
  7. Choose I have read and agreed to <provider>‘s model terms. If the provider doesn’t supply a terms URL, choose I agree to <provider>‘s model terms instead.
  8. Choose Next: Add Model API.
Baseten adds the Model API to the workspace and opens its details page. Developers in that workspace can call it with their own Baseten API keys.

​
Give customers access to a Dedicated model

Baseten configures Dedicated customer availability during assisted setup. Dedicated models do not use activation URLs. Each customer deployment remains pinned to the version the customer adopts.

​
Set Model API rate limits

Distribution Model APIs support per-customer requests-per-minute (RPM) and tokens-per-minute (TPM) limits. Manage these limits in the console. Baseten doesn’t expose a supported public API for this operation. To set RPM and TPM limits:
  1. Sign in to your workspace at app.baseten.co.
  2. Choose Insights under Labs in the sidebar.
  3. In the Customers table, choose the customer row.
  4. Under Model APIs, choose the ⋯ menu for the Model API.
  5. Choose Edit limits.
  6. For Requests per minute, type the RPM limit.
  7. For Tokens per minute, type the TPM limit.
  8. Choose Save.
The console confirms the update with Updated limits for <model name>.

​
What stays private

You see named customers and their Distribution channels. Prompts, outputs, and request-level logs stay inside the customer’s Baseten boundary.

​
Next steps