Support fast/extended-thinking mode flag #1

Open
opened 2026-05-23 10:09:19 +00:00 by fkrebs · 0 comments
Owner

Request

LiteLLM has a /fast toggle concept (Claude Opus with faster output). The claude-max-bridge should support a way to request fast mode or extended thinking via the API.

Possible approach

  • Accept extra_body.fast: true → pass --fast flag to Claude CLI (if supported)
  • Or map model suffix -fast to the appropriate CLI flags
  • Check if claude --fast is a valid CLI option first: claude --help | grep fast

Notes

Claude Code fast mode uses Claude Opus with faster output. The CLI may support this via a flag or it may be a UI-only feature. Verify before implementing.

## Request LiteLLM has a `/fast` toggle concept (Claude Opus with faster output). The claude-max-bridge should support a way to request fast mode or extended thinking via the API. ## Possible approach - Accept `extra_body.fast: true` → pass `--fast` flag to Claude CLI (if supported) - Or map model suffix `-fast` to the appropriate CLI flags - Check if `claude --fast` is a valid CLI option first: `claude --help | grep fast` ## Notes Claude Code fast mode uses Claude Opus with faster output. The CLI may support this via a flag or it may be a UI-only feature. Verify before implementing.
Sign in to join this conversation.
No Label
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: fkrebs/claude-max-bridge#1