Mistral AI · Tools

Mistral Le Chat

Visit website

At a glance

Price
from 14.99 $/mo
Vendor
Mistral AI

Specifications & properties

Key decision factors

Pricing model
Freemium 1
Pricing model
Freemium 1
Context window
256,000 tokens 1
Citation / source disclosure
Inline citations 1
Input price per 1M tokens
0.5 USD per 1M tokens 1
GDPR / EU hosting
EU hosting 1
GDPR / EU hosting
EU hosting 1
Real-time web
Yes 1
Data privacy
Opt-out available 1
Open weights / self-hosting
Yes 1

Pricing

Price from
14.99 $/mo 1
Price from
14.99 $/mo 1
Free tier
Yes 1
Free tier
Yes 1
Output price per 1M tokens
1.5 USD per 1M tokens 1

Integration

API available
Yes 1
API available
Yes 1

Data

Index source
Web search 1

Performance

Output speed
43.2 tokens/s 2

Model

Underlying model
Mistral AI's own models (Mistral Large, Mistral Medium 3.5, Mistral Small) 1

Capabilities

Team collaboration
Yes 1

Compliance

SOC 2
Yes 1
Report data / suggest a correction

Metrics vs. the category

Context window
256,000 tokens Category max 1,050,000 tokens (6 with data)
Input price per 1M tokens
0.5 USD per 1M tokens Category max 10 USD per 1M tokens (7 with data)
Output price per 1M tokens
1.5 USD per 1M tokens Category max 50 USD per 1M tokens (7 with data)
Output speed
43.2 tokens/s

Le Chat (now branded Mistral Vibe) is Mistral AI's assistant and answer engine that combines the company's own models with real-time web search to deliver evidence-based answers with citations. The Paris-based vendor hosts user data in the European Union by default.

Profile

Mistral Le Chat (now branded Mistral Vibe) is Mistral AI's conversational assistant and answer engine, built by the Paris-based Mistral AI, that combines the company's own models with real-time web search to deliver evidence-based, cited answers. It runs on a freemium model with a genuine free tier, and the Pro plan starts at $14.99 per month for long-running tasks and all-day coding access.

Who builds it

Le Chat is developed by Mistral AI, a Paris, France-based AI lab. By default, user data is hosted in the European Union, with an optional US API endpoint for those who explicitly choose to route data there instead.

What it does

Le Chat combines the pre-trained knowledge of Mistral's own models with recent information balanced across web search, journalism, and other sources, producing well-reasoned, evidence-based answers with detailed citations. For most tasks it defaults to Mistral Large; Mistral Medium 3.5 is used for coding, and Mistral Small for cost-sensitive workloads. Mistral Large 3 runs on a 256,000-token context window and is released as an open-weight model under the Apache 2.0 license, alongside API pricing of $0.50 per 1M input tokens and $1.50 per 1M output tokens.

Core features

  • Freemium pricing: a genuine free tier, with Pro starting at $14.99/month and Team at $24.99/user/month
  • Real-time web search with inline, detailed citations for evidence-based answers
  • Underlying models: Mistral Large, Mistral Medium 3.5 (coding), and Mistral Small (cost-sensitive tasks)
  • Mistral Large 3 is released open-weight under the Apache 2.0 license
  • 256,000-token context window; API priced at $0.50/1M input and $1.50/1M output tokens
  • EU data hosting by default, with an optional US API endpoint
  • SOC 2 Type II and ISO 27001/27701 compliance
  • Users can opt out of having their data used for model training on Free, Pro, and Team plans

Who it's for

Le Chat suits individuals and teams who want a European-hosted alternative to US-based assistants, along with developers who want an open-weight model family they can also self-host or fine-tune. The free tier covers everyday use with limited messages and web searches; Pro removes those limits for long-running tasks and coding, and Team adds a secure collaborative workspace for organizations building with AI together.

Data and compliance

By default, Mistral hosts Le Chat user data in the European Union; users can explicitly opt into a US API endpoint if they need data processed there instead. Mistral states compliance with both SOC 2 Type II and ISO 27001/27701 frameworks, and its plan comparison shows an opt-out control for model training available across the Free, Pro, and Team tiers — so data isn't used to train Mistral's models unless a user leaves that setting on its default.

Frequently asked questions

What does Mistral Le Chat cost?

Mistral Le Chat's Pro plan starts at $14.99 per month, giving full access to Vibe for long-running tasks and all-day coding; a Team plan for collaborative workspaces costs $24.99 per user per month. A free plan is also available with limited messages and web searches.

Is there a free version of Mistral Le Chat?

Yes. Mistral offers a free plan with access to Mistral's latest models, chat, search, image creation, and Vibe for agentic coding in the terminal, though with limited messages and web searches compared to the paid Pro plan.

Does Mistral Le Chat have an API?

Yes. Mistral's API is billed per million tokens processed for both input and output — for example, Mistral Large 3 costs $0.50 per 1M input tokens and $1.50 per 1M output tokens — separate from the Le Chat consumer subscription.

What can you use Mistral Le Chat for?

Le Chat is a general-purpose assistant that combines Mistral's own models with real-time web search, journalism, and other sources to produce evidence-based answers with detailed citations, and includes agentic coding support through Vibe for code in the terminal.

Mistral Le Chat vs ChatGPT — which should I pick?

Both are freemium assistants with real-time web access and a paid tier under $20/month. Mistral Le Chat starts at $14.99/month, runs on Mistral's own open-weight models with a 256,000-token context window, and hosts data in the EU by default; ChatGPT (OpenAI) starts at $8/month, defaults to GPT-5.6 with a ~1.05M-token context window, and doesn't host data in the EU by default. Pick Le Chat for EU data hosting and open-weight models, ChatGPT for the larger context window and broader ecosystem.

Is Mistral Le Chat GDPR-compliant, and does it train on my data?

By default, Mistral hosts Le Chat user data in the European Union — an explicit US API endpoint is available for those who choose to route data there instead. Mistral states SOC 2 Type II and ISO 27001/27701 compliance, and its Free, Pro, and Team plans all show an opt-out control for model training, so your data isn't used to train Mistral's models unless you leave that setting on its default.