llm-mistral 0.16 Adds Reasoning Model Support with Mistral Large 4

It is a plugin for Simon Willison's llm tool that lets you run and call Mistral models from the command line or Python.

It adds support for reasoning models, including Mistral Large 4, which the source describes as newly released.

Reasoning models generate intermediate thinking steps before the final answer, which helps with multi-step tasks like analysis and planning, whereas standard models answer directly.

If you already use llm with Mistral models, the update widens your options without changing your workflow. If you do not use the tool, there is no urgent reason to move.