Guides
Long sessions
Keep /compact working on any model
Claude Code's /compact asks the model for a plain-text summary. If tools are still offered, a non-Anthropic model may answer with a tool call, and the summary step rejects it.
What OpenLLM does
On a detected compaction turn routed to a non-Anthropic model, OpenLLM drops the tool list and sets tool_choice to none, so the model can only answer with text.
What it leaves alone
Requests to Anthropic itself pass through unchanged — Claude Code's own compaction already works there.
Fallback chains when a model is rate-limited
Define an ordered chain per model alias. When a hop returns a rate limit, an overload or a provider error, the request moves to the next hop instead of failing your session.
- Chains are configured once and shared by every client and device
- Each hop resolves its own model settings, so a fallback never inherits the wrong parameters
- Every hop that ran is visible in your request history
Questions
Something else? Contact us