Mistral AI
Mistral AI is a French AI company offering open-weight and frontier models, the Studio developer platform and API, Forge for customisation and its Vibe assistant.
About this tool
Mistral AI builds language models and the software to deploy them, with an emphasis on letting organisations keep control. Its catalogue includes frontier models and specialist open-weight models for text, code, vision and audio, which companies can use as they are or adapt with Forge, its customisation offering. Models can run on-premises, in a customer’s virtual private cloud or on Mistral’s infrastructure.
For developers, Studio is the console and API: API keys, a playground, evaluations, agents and SDKs. For end users, Mistral’s assistant now appears under the name Vibe, which the pricing page links to Le Chat, offering chat on web and mobile plus coding through a CLI and IDE plugins for VS Code and JetBrains powered by its Devstral models.
What you can do
Open-weight and frontier models
A range of models for text, coding, multimodal and audio tasks, many available with open weights.
Studio and API
A developer console with keys, playground, evaluations, agent building and SDKs for production apps.
Forge customisation
Tailor models to an organisation’s data, use cases and requirements.
Vibe assistant
An assistant for research and everyday tasks, with a coding agent in the terminal and IDE.
Who is it for?
- A European company deploys a Mistral model inside its own cloud to keep data in-region.
- A developer builds a document search agent with the Mistral API and RAG tooling.
- An engineering team uses the Vibe CLI for autonomous coding tasks.
Before you choose
Product names have changed recently, with the assistant now presented as Vibe; check documentation for current naming before purchasing.
Open-weight licences differ by model; review the licence of each model before commercial deployment.
Frequently asked questions
Is Le Chat still available?
Mistral’s pricing now refers to its assistant as Vibe and mentions Le Chat in parentheses, suggesting the assistant was renamed. Check Mistral’s site for details.
Can Mistral models be self-hosted?
Yes. Mistral offers open-weight models and deployment on-premises or in your VPC, as well as hosted options.
How much does Mistral cost?
There is a free tier for the assistant, Pro at $14.99 per month, team and enterprise plans, and per-token API pricing by model.
Check the official website for current features, pricing, and terms.
Similar tools
Mistral Vibe (Le Chat)
ProductivityVibe is Mistral AI’s assistant, previously known as Le Chat, for research, writing and everyday tasks, with a coding agent available on web, terminal, IDE and mobile.
LM Studio
CodingLM Studio is a desktop app for discovering and running open AI models locally, with an agent called Bionic, developer APIs and local voice transcription.
DeepSeek
CodingDeepSeek develops large language models, including open-weight reasoning models, offered through a chat app and an OpenAI-compatible developer API.
Jan
CodingJan is a free, open-source desktop app for chatting with AI models that run offline on your computer, with optional connections to cloud models.
Google AI Studio
CodingGoogle AI Studio is Google’s web workspace for testing Gemini models, getting a Gemini API key, building prototypes and moving them into production.
Weights & Biases
CodingWeights & Biases is an AI developer platform for tracking experiments, evaluating LLM apps and agents with Weave, and managing models; it is now part of CoreWeave Forge.