This is harder to build than it looks. Preserving full conversational context while swapping underlying model providers mid-flight is a surp…

This is harder to build than it looks. Preserving full conversational context while swapping underlying model providers mid-flight is a surprisingly deep systems problem. Most tools drop state or force you to start over. deepagents-cli does this natively: swap models mid-conversation with zero context loss. Try it: https://docs.langchain.com/oss/python/deepagents/cli/overview I love how you can just switch to any model provider using preexisting context with @cursor_ai. This is a super power for me. Was stuck in a bug and an ugly UI design. Switched to another model (using the whole existing context) and it solved it in a single shot.

Source: Harrison Chase (X) | 2026-05-11