Building the Claude API for the Frontier

Every Claude launch puts the spotlight on the model, but just as much work goes into making sure the API evolves with it. That means adding new controls, retiring outdated ones, and helping developers rethink how they build with each new release.

Jake Dejno (Member of Technical Staff, Claude Platform) and Lucca Siaudzionis (Member of Technical Staff, Platform Intelligence) discuss how research and customer requests shape what ships, best practices for changing effort mid-conversation without breaking prompt caching, and how Mythos and Fable use the same model weights with different safeguards.**

0:00 - Intro
0:15 - How the Messages API works
1:40 - Deciding what features to release
4:15 - Handling model versioning and migrations
8:15 - Best practices for managing cost
11:22 - Balancing access and safety with Mythos and Fable
14:41 - What’s next for the API

Learn more about the Claude Platform

Follow ClaudeDevs on X for product updates and best practices from the Claude Platform team
*Claude Managed Agents and changing effort mid-conversation are in beta.
**Claude Mythos access is through the Cyber Verification Program.
33

Comments 0

Replying to
5000 max

Be the first to comment!

This website uses cookies
This website uses cookies.
To maintain our services, we use cookies. They allow us to personalize ads, analyze our traffic, and provide social media features.

Customize your preferences

Select the cookies you accept: