reasoning

Add chain-of-thought scaffolding to each message, helping smaller local models reason before they answer. A lightweight prompt-preprocessor — no tools, no setup.

Part of lmstudio-suite.

Configuration (per-chat)

• Reasoning mode:
  • Off — pass the message through unchanged.
  • Concise (default) — "think step by step, then give a concise final answer".
  • Full — explicit reasoning followed by a marked Final answer: line.

Use

Enable it in a chat and pick a mode. It appends the reasoning instruction to your message before the model sees it.

Note: this shapes the model's input. For hard schema/JSON enforcement on the output, use generateStructured from @lmstudio-suite/core in a standalone SDK app.

MIT licensed.