Skip to content
Back

AI posts

AI release notes

Mistral AI's official changelog: new Mistral models and changes for developers.

Compiled by the Toolstash team · Updated

  1. New model

    Z.ai GLM 5.3 Flash is now available

    A multimodal 320B-parameter (18B active) Mixture-of-Experts model from Z.ai with a 1M-token context window. Built for developer agent workflows, visual coding, and complex tool-use tasks. Pricing: $0.15/M input tokens, $0.03/M cached input, $0.50/M output tokens.

  2. Mistral APIAnnouncement

    We released Voice Agents, a managed real-time bidirectional voice pipeline exposed through the Voice Agents…

    We released Voice Agents, a managed real-time bidirectional voice pipeline (microphone capture, STT, LLM, TTS, playback) exposed through the Voice Agents SDK, with telephony, browser, and custom-audio interfaces.

  3. New modelAnnouncement

    We released the Voxtral 3 audio family, our next generation of speech-to-text and text-to-speech models

    All four models are in Public Preview. Voxtral TTS 3 (voxtral-mini-tts-3-0): text-to-speech with zero-shot voice cloning, streaming, and contextual delivery. Voxtral Mini Transcribe 3 (voxtral-mini-transcribe-3-0): offline transcription and diarization with timestamps. Voxtral Mini Transcribe Realtime 3 (voxtral-mini-transcribe-realtime-3-0): real-time transcription. Voxtral Small Transcribe 3 (voxtral-small-transcribe-3-0): offline transcription and diarization with timestamps.

  4. New model

    Mistral Large 4 is now available in Public Preview

    For details, see the Mistral Large 4 model card. Mistral Large 4 is an open-weight, general-purpose multimodal model with a Mixture-of-Experts architecture and a 1M context window. Open weights are coming soon. Launch pricing: 50% off for 2 weeks.

  5. Mistral API

    OCR 4.0 is deprecated and retires on September 30, 2026

    Use OCR 4.1 (mistral-ocr-4-1 or mistral-ocr-latest) instead, at the same price. Leanstral 1.5 (labs-leanstral-1-5), an experimental Labs model, is deprecated and retires on September 30, 2026. Z.ai GLM 5.2 (zai-glm-5-2) is deprecated and retires on October 31, 2026.

  6. New model

    Z.ai GLM 5.3 is now Generally Available

    Z.ai GLM 5.3 (zai-glm-5-3) is now Generally Available.

  7. New model

    OCR 4.1 is now Generally Available

    OCR 4.1 (mistral-ocr-4-1) is now Generally Available.

  8. Mistral API

    The OCR API confidence_scores_granularity parameter now supports "block" granularity

    "page" returns page-level scores only, "block" returns page-level and block-level scores, and "word" returns page-level and word-level scores.

  9. New model

    We released OCR 4.1. mistral-ocr-latest and mistral-ocr-4 now point to it

    We released OCR 4.1 (mistral-ocr-4-1). mistral-ocr-latest and mistral-ocr-4 now point to it.

  10. New model

    We released Leanstral 1.5, an updated Lean 4 formal proof engineering model with improved SFT mixture…

    This model will be retired on September 30, 2026.

  11. Mistral API

    Introducing include_blocks in our OCR API

    When set to true, each page returns a blocks array with paragraph-level bounding boxes and a structural label (text, title, list, table, image, equation, caption, code, references, aside_text, header, footer, signature) in reading order. Learn more in our OCR documentation. The pages parameter in our OCR API now also accepts a string of comma-separated digits and ranges (e.g.

  12. New model

    We released OCR 4. mistral-ocr-latest now points to it

    We released OCR 4 (mistral-ocr-4-0). mistral-ocr-latest now points to it.

  13. Mistral API

    We launched Vibe, our unified agent at chat.mistral.ai, available in three modes:

    Work for productivity on web and mobile, with Skills, Workflows, Connectors, Libraries, and scheduled tasks. Code for developers, covering the Vibe CLI, the VS Code extension, and Vibe Code Web for remote coding sessions in a managed cloud sandbox. Chat preserves the legacy Le Chat experience for existing workflows. Documentation restructured: previous /le-chat/* and /mistral-vibe/* paths now redirect to the new /vibe/* tree.

  14. New model

    We released Mistral Medium 3.5, our frontier-class multimodal model optimized for agentic and coding use…

    Released as open weights under a Modified MIT license.

  15. New model

    We released Voxtral TTS, our state-of-the-art text-to-speech model with zero-shot voice cloning,…

    We released Voxtral TTS (voxtral-tts-2603), our state-of-the-art text-to-speech model with zero-shot voice cloning, multilingual support, and real-time streaming.

  16. New model

    We released Mistral Small 4, a hybrid model unifying instruct, reasoning, and coding in a single multimodal…

    We released Leanstral (labs-leanstral-2603), our first open-source code agent designed for Lean 4 formal proof engineering.

  17. Mistral API

    We added Custom Guardrails support for Agents and Conversations

    Guardrails can now be configured directly on an Agent via the guardrails parameter. Guardrails can be passed per-request on POST /v1/conversations using the guardrails field. Guardrails can be passed per-request on POST /v1/chat/completions using the guardrails field.

  18. New model

    We released Mistral Moderation 2603

    We released Mistral Moderation 2603 (mistral-moderation-2603).

  19. New model

    We released Voxtral Mini Transcribe 2 and Voxtral Mini Transcribe Realtime

    Introducing context biasing in our Audio Transcriptions API, Introducing diarize in our Audio Transcriptions API.

  20. Mistral API

    We released Vibe 2.0

    Document Annotations update and improvements, 8 page limit removed.

  21. New model

    Devstral 2.0 moves to paid API access

    Devstral 2.0 moves to paid API access.

  22. Mistral API

    We released inline batching allowing the creation of batch jobs without file uploading

    We released inline batching allowing the creation of batch jobs without file uploading.

  23. Mistral API

    Introducing table_format in our OCR API, allowing you to choose between markdown and html for table formatting

    Introducing extract_footer, extract_header in our OCR API, as well as hyperlinks output.

  24. New model

    We released OCR 3

    We released OCR 3 (mistral-ocr-2512).

  25. New model

    We released Mistral Small Creative as a Labs model

    We released Mistral Small Creative (labs-mistral-small-creative) as a Labs model.

  26. Mistral API

    We released Mistral Vibe

    We released Mistral Vibe.

  27. New model

    We released Devstral 2 and Devstral Small 2

    We released Devstral 2 (devstral-2512) and Devstral Small 2 (labs-devstral-small-2512).

  28. New model

    We released Mistral Large 3 and Ministral 3

    We released Mistral Large 3 (mistral-large-2512) and Ministral 3 (ministral-3b-2512, ministral-8b-2512 and ministral-14b-2512).

  29. New model

    We released Magistral Medium 1.2 and Magistral Small 1.2

    We released Magistral Medium 1.2 (magistral-medium-2509) and Magistral Small 1.2 (magistral-small-2509).

  30. Mistral API

    Added a new parameter p to the chunks streamed back by the Completion API

    Implemented for security to prevent token-length side-channel attacks, as reported by Microsoft researchers. Note that this change may break applications relying on strict parsing of the chunks.