Mistral API
Add AI to your app with Mistral AI.
Access Mistral AI models through a single API.
AI Chat API
Multimodal Understanding
Processes both text and images natively, with built-in vision capabilities for document analysis and visual reasoning.
Multilingual by Design
Native support for 40+ languages, with strong performance beyond just English and Chinese.
Chain-of-Thought Reasoning
Dedicated reasoning models for complex logic, math, and analytical tasks.
Agentic & Tool Use
Supports tool calling for automated workflows, connecting to external APIs and data sources.
256K Context Window
Handle massive documents, long conversations, and complex multi-step tasks.
Specialized Code Models
Devstral models built specifically for coding tasks, bug fixing, and multi-file codebase navigation.
Mistral Models
Access all Mistral AI models through a single API.
Mistral Large 3
Flagship model with 256K context window for complex reasoning and enterprise tasks.
mistralai/mistral-large-2512
Mistral Medium 3.1
Balanced model for general tasks with excellent performance-to-cost ratio.
mistralai/mistral-medium-3.1
Mistral Small 3.2
Fast and efficient 24B model optimized for quick responses and lower latency.
mistralai/mistral-small-3.2-24b-instruct
Mistral Small Creative
Optimized for storytelling, creative writing, and imaginative content generation.
mistralai/mistral-small-creative
Codestral
Specialized model for code generation, completion, and programming tasks.
mistralai/codestral-2508
Devstral
State-of-the-art open model for agentic coding, bug fixing, and codebase navigation.
mistralai/devstral-2512
Devstral Medium
Medium-sized agentic coding model for balanced performance and efficiency.
mistralai/devstral-medium
Devstral Small
Compact agentic coding model for fast code assistance and lightweight deployments.
mistralai/devstral-small
Pixtral Large
Large multimodal model with vision capabilities for image understanding and analysis.
mistralai/pixtral-large-2411
Pixtral 12B
Efficient 12B multimodal model for image-to-text and visual reasoning tasks.
mistralai/pixtral-12b
Mixtral 8x22B
Mixture-of-experts model with 8 experts at 22B each for high-quality outputs.
mistralai/mixtral-8x22b-instruct
Mixtral 8x7B
Efficient mixture-of-experts model with 8 experts at 7B for fast, quality responses.
mistralai/mixtral-8x7b-instruct
Mistral Nemo
Lightweight model optimized for speed and efficiency in resource-constrained environments.
mistralai/mistral-nemo
Ministral 8B
Compact 8B model for everyday tasks with good quality-to-speed balance.
mistralai/ministral-8b
Ministral 3B
Ultra-lightweight 3B model for edge deployment and fast inference.
mistralai/ministral-3b
Voxtral Small
Speech and audio-capable model for voice interaction and transcription tasks.
mistralai/voxtral-small-24b-2507
Add AI Chat in Seconds
Add Mistral AI to your app with just a few lines of code.
No API keys, no backend, no configuration required.
<html>
<body>
<script src="https://js.puter.com/v2/"></script>
<script>
puter.ai.chat("Explain machine learning to a beginner", {
model: "mistralai/mistral-large-2512"
}).then(response => {
document.body.innerHTML = response.message.content;
});
</script>
</body>
</html>
Related Resources
Learn how to use Mistral and other AI models with Puter.js.