General purpose
Models built for reasoning, language, multimodal understanding, and coding.
Specialist
Models built to excel at specific tasks and domains, including document intelligence, audio, robotics, and physical AI.
Publications.
Discover how our models and methods advance reasoning, perception and action.
Filter by topic
19 results
Shieldstral
Policy-adaptive multimodal safety classification.
Safety
Robostral Navigate
Vision-only navigation across robot platforms.
Robotics
Phoenix-VL 1.5 Medium
Localized multimodal models without capability loss.
Applied AI
Voxtral TTS
Expressive multilingual speech generation and adaptation.
Audio
Voxtral Realtime
Sub-second streaming speech recognition.
Audio
Ministral 3
Cascade-distilled models for constrained compute.
Foundation models
Research areas.
Safety and guardrailing.
Simulation.
Inside the work.
Go deeper on the techniques, systems, and engineering behind our models—from training and evaluation to inference and efficiency.
Research
Introducing Mistral Large 4
Research
Introducing Mistral OCR 4
State of the art document intelligence model.
Research
Introducing Robostral Navigate
Robostral Navigate, our first model built for embodied navigation.
Research
Leanstral 1.5: Proof Abundance for All
Solutions
Introducing Shieldstral.
Research
Physics AI research that’s shaping the industry.
Published breakthroughs pushing the state of the art.
Research
Speaking of Voxtral
Voxtral TTS: A frontier, open-weights text-to-speech model that’s fast, instantly adaptable, and produces lifelike speech for voice agents.
Build on our open-weight models.
Download, adapt, and deploy our models.
