The Distribution Platform is part of Baseten for Model Labs. Baseten enables it per workspace. To get access, talk to us.
Choose a distribution channel
Choose the channel that matches how customers should consume your model. You can publish the same model through both channels.Model API
Model APIs use a shared endpoint and can serve models across modalities and runtimes. Each Model API exposes the request and response interface that the lab defines.Dedicated deployment
Dedicated gives each customer an isolated deployment pinned to a published version of your model. This is useful when customers need isolated infrastructure.How distribution works
To distribute a model through Baseten:- Deploy: Use Manage your model to deploy to your Baseten workspace and save the model and model version IDs. For supported deployment workflows, see Deploy and iterate.
- Validate: Use Manage your model to send representative requests to the deployed model and check its responses. For request formats and authentication, see Call your model.
- Publish: Use Manage your
model to publish through one or
both distribution channels:
- Create a Dedicated listing for customer-isolated deployments.
- Create a Model API for a shared endpoint.
- Update: Use Manage your
model when you deploy a
new version:
- Publish a new Dedicated listing version, make it live for adoption, and have the customer adopt it.
- Point the Model API at the new model.
- Retire: Use Manage your
model when you no longer want to
distribute it:
- Delete the Dedicated listing.
- Delete the Model API endpoint.
Dedicated version adoption
Each customer deployment stays pinned to the listing version that the customer adopted. Making a newer listing version live changes the version available for new adoption, but it does not update existing customer deployments. A customer adopts the newer version when they are ready; the older deployment continues to run until the customer replaces or retires it. Deleting a Dedicated listing stops new adoption and removes its catalog resources. It does not terminate customer deployments already created from an adopted version. Those deployments continue until each customer replaces or retires them.Customer availability and artifact access
A lab can start a Model API with an allowlist of selected customer workspaces, then make it available to all eligible Baseten customers after validation. Baseten configures Model API allowlists and broader availability during assisted setup. For Dedicated listings,is_public controls whether the listing is
publicly accessible.
closed_source and allow_truss_download control artifact access, not customer
availability. Closed-source Dedicated listings prevent customers from viewing
or downloading the Truss and weights. The Baseten Delivery
Network delivers model weights to customer deployments
without exposing source credentials, and Baseten blocks API operations that
expose artifacts. For Dedicated deployments, customers cannot view the Truss
configuration, download artifacts, run truss push,
use SSH, or change the instance type. Set closed_source when you create a
Dedicated listing and keep allow_truss_download set to false for each
published version.
Next steps
- Manage your model: Create a Dedicated listing or Model API endpoint and complete the assisted setup.
- Report usage: Report billable usage that the gateway cannot meter.
- Give customers access: Understand customer activation and limits.
- Insights: Track customers, models, and proceeds.