Introducing canary deployments for seamless promotions

We're excited to introduce canary deployments on Baseten, designed to phase in new deployments with minimal impact on production latency and uptime.

When enabled for a model, Baseten gradually shifts traffic to the newly promoted deployment over a configurable window, ramping up in 10 equal steps. This ensures that even at peak traffic, the new deployment scales appropriately as traffic shifts. You can enable and configure settings in the UI under "Configure promotion" or via the REST API.

Check out our launch post for more information!

✕