# What changed in langchain-mistralai

> 16 changes in langchain-mistralai 1 from its official release notes, upgrade guides and advisories: 1 breaking, 8 new APIs and options. Each one cited. Sources checked 2026-11-12. HTML: https://unstale.sh/pypi/langchain-mistralai/

To see which of these apply to a repository, run `npx @unstale/cli init` in it.

## langchain-mistralai 1

### Breaking changes

- Tenacity retries are removed for embeddings. (1.0.1) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.0.1)

### New APIs and options

- Chat responses now surface citation metadata, allowing developers to access citations returned with responses. (1.1.6) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.6)
- Mistral chat models now support stop sequences. (1.1.6) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.6)
- Adds content-block-centric streaming (v2). (1.1.3) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.3)
- Human messages now support image input. (1.1.3) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.3)
- Model profiles now include text_inputs, exposing text input information for models. (1.1.2) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.2)
- Model profiles now include text_outputs, exposing text output information for models. (1.1.2) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.2)
- Adds support for the reasoning feature. (1.0.0) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.0.0)
- Adds support for v1 content. (1.0.0) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.0.0)

### Fixes and other changes

- Callbacks now support content block tokens. (1.1.5) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.5)
- Non-wire keys are stripped from ToolMessage before sending. (1.1.4) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.4)
- Exception retry logic in embeddings was updated. (1.1.2) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.2)
- Fixed a NameError when RunnableWithFallbacks.__getattr__ resolves type hints. (1.1.2) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.2)
- Handles null content in tool call responses. (1.1.1) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.1.1)
- Fixes duplicate tool calls when converting to a Mistral chat message. (1.0.1) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.0.1)
- Handles HTTP errors in async embed documents. (1.0.0) — [release notes](https://github.com/langchain-ai/langchain/releases/tag/langchain-mistralai%3D%3D1.0.0)