Mistral Pricing & Le Chat Plans: Complete Guide (2026)
Mistral AI structures its pricing around two products: Le Chat (the consumer and business chatbot) and the Mistral API (for developers building on Mistral's large language models). As of May 2026, Le Chat offers four tiers -- Free, Pro, Team, and Enterprise -- while the API follows a pay-per-token model with rates that vary by model size and capability.
This guide walks through every tier, explains what you actually get at each price point, covers the full API rate card, and helps you decide which plan fits your workflow. All prices are in USD and exclude applicable taxes.
What Is Mistral Pricing?
Mistral splits its offerings into two billing models. Le Chat is a subscription product with monthly tiers, similar to ChatGPT Plus or Claude Pro. The Mistral API (hosted on La Plateforme) is a pay-per-use service where you are billed per million tokens processed, with separate rates for input and output.
The two products serve different audiences. Le Chat targets end users who want a chat interface for daily work -- writing, research, coding, and image generation. The API targets developers and businesses embedding Mistral models into their own applications, pipelines, or products.
Le Chat Free: $0/Month
The Free plan is a genuine starting point, not a crippled trial. You get access to Mistral's state-of-the-art models, core chat features, Flash Answers for quick lookups, a code interpreter, document uploads, and image generation. The plan also includes 40+ enterprise connectors (GitHub, Asana, Snowflake, and others) and the ability to save up to 500 memories.
The main constraint is usage volume. Mistral applies a soft cap of roughly 25 messages (or equivalent usage) before features slow down or pause for the day. Resource-intensive tasks like image generation and deep research consume that cap faster than simple text queries. There is no dedicated support on the Free tier.
What You Get
- Access to Mistral's SOTA AI models
- Chat, web search, Think mode (preview), deep research (preview)
- Image generation and code interpreter
- Canvas mode for collaborative editing
- 40+ enterprise connectors (GitHub, Asana, Snowflake, Outlook, Stripe, Zapier)
- Up to 500 memories for personalized context
- Group chats into projects
Best for: Students, hobbyists, and anyone evaluating whether Mistral fits their workflow before committing to a paid plan.
Le Chat Pro: $14.99/Month
The Pro plan is Mistral's most popular paid tier for individual users. At $14.99 per month (or $6.99 per month with a verified student discount), it significantly expands what you can do compared to Free.
Usage Multipliers vs. Free
| Feature | Free | Pro (multiplier) |
|---|---|---|
| Messages | Base (~25/day) | Up to 6x (~150/day) |
| Web searches | Base | Up to 5x |
| Think mode | Base | Up to 30x |
| Deep research | Base | Up to 5x |
| Image generation | Base | Up to 40x |
| Document uploads | Base | Up to 20x |
| Code interpreter | Base | Up to 5x |
| Memories | 500 | 1,000 |
| Document storage | Limited | Up to 15 GB |
| Projects | Limited | Up to 1,000 |
Pro-Only Features
- No Telemetry Mode: Toggle your account so none of your prompts or data are used to train Mistral's models
- 150 Flash Answers per day: Ultra-fast brief answers for quick facts
- Mistral Vibe: All-day coding workspace with pay-as-you-go billing beyond included limits
- Chat and email support
- Full connectors directory access (beta)
The No Telemetry Mode is a standout feature. When enabled, Mistral guarantees that nothing you type is retained or used for model training. This matters if you work with sensitive business content or proprietary code.
Best for: Power users, freelancers, solo developers, and professionals who rely on AI daily but do not need team collaboration features.
Le Chat Team: $24.99/User/Month
The Team plan is built for collaborative workspaces where multiple people need shared access to Mistral's tools. It costs $24.99 per user per month on monthly billing, dropping to $19.99 per user per month when billed annually.
What Team Adds Over Pro
- 30 GB storage per user (double the Pro limit)
- Domain name verification for organizational identity
- Data export capabilities
- Centralized billing and admin controls
- Shared knowledge and RAG library for team-wide document access
- Opted out of data sharing by default with zero data retention options
Every Team user gets the same usage multipliers as Pro (6x messages, 30x Think mode, 40x image generation). The difference is organizational control: admins can manage seats, enforce data policies, and maintain a shared knowledge base that everyone on the team can query.
Best for: SMEs and startups with 5 to 20 users who need collaboration tools, shared libraries, and stronger data governance.
Le Chat Enterprise: Custom Pricing
Enterprise pricing is custom and quote-based, with typical entry points starting around $20,000 or more per month. Organizations negotiate directly with Mistral for a tailored package that depends on user count, usage volume, deployment model, and support level.
Enterprise-Only Features
- Private or on-premise deployment (self-hosted, private cloud, or Mistral-hosted)
- Custom models fine-tuned on organizational data
- Custom UI and tools including white-label options
- No-code agent builders for workflow automation
- Full audit logs and compliance reporting
- SAML SSO for identity management
- Admin API for programmatic user and policy management
- Voice mode and async coding agents
- Custom MCP connectors for proprietary integrations
- Dedicated help center and knowledge base
Le Chat Enterprise launched in May 2025 and is available on Google Cloud Marketplace, with Azure AI and AWS Marketplace availability planned. Organizations like HSBC and BNP Paribas have adopted Mistral's enterprise tier for privacy-sensitive banking workflows.
Best for: Large organizations, regulated industries (finance, healthcare, government), and any business requiring on-premise deployment or custom model training.
API Pricing by Model
Mistral's API charges per 1 million tokens processed, with separate rates for input and output. Prices vary by model capability, and cache-read discounts are available on most models.
Flagship Models
| Model | Context | Input / 1M | Output / 1M |
|---|---|---|---|
| Mistral Large 3 (2512) | 262K | $0.50 | $1.50 |
| Mistral Medium 3.1 | 131K | $0.40 | $2.00 |
| Mistral Medium 3 | 131K | $0.40 | $2.00 |
| Magistral Medium | 40K | $2.00 | $5.00 |
Small and Efficient Models
| Model | Context | Input / 1M | Output / 1M |
|---|---|---|---|
| Mistral Small 3.2 24B | 131K | $0.075 | $0.20 |
| Mistral Small 3.1 24B | 128K | $0.10 | $0.30 |
| Mistral Small 24B 2501 | 33K | $0.05 | $0.08 |
| Saba | 33K | $0.20 | $0.60 |
Specialized Models
| Model | Context | Input / 1M | Output / 1M |
|---|---|---|---|
| Codestral 2508 | 256K | $0.30 | $0.90 |
| Pixtral Large 2411 | 131K | $2.00 | $6.00 |
| Pixtral 12B | 128K | $0.10 | $0.10 |
| Devstral Small 2505 | 128K | $0.00 | $0.00 |
Devstral Small 2505 is currently free to use through the API, making it an accessible entry point for developers exploring Mistral's code-focused models.
How billing works: You pay only for tokens consumed. Input tokens (your prompts) and output tokens (the model's response) are billed separately. Most models also offer a cache-read discount (typically 10x cheaper than standard input) when the same prompt prefix is reused across requests.
Mistral Forge Platform
Announced at NVIDIA GTC in March 2026, Mistral Forge is an enterprise platform for building custom frontier-grade AI models on proprietary data. Unlike standard fine-tuning or RAG approaches, Forge supports full pre-training, post-training, and reinforcement learning on internal datasets.
Forge uses a software license fee model rather than charging for compute. Organizations run training workloads on their own GPU clusters and pay Mistral for the platform tooling. Optional "forward-deployed scientist" support is available for teams that need hands-on guidance.
Early adopters include ASML (semiconductor lithography), Ericsson (telecom), the European Space Agency, and Singapore's DSO National Laboratories. Forge pricing is negotiated per contract and is not published on Mistral's standard pricing page.
Competitor Comparison
How does Mistral's pricing compare to other major AI platforms as of May 2026?
| Feature | Mistral Pro | ChatGPT Plus | Claude Pro | Gemini Advanced |
|---|---|---|---|---|
| Monthly price | $14.99 | $20 | $20 | $19.99 |
| Student discount | $6.99/mo | None | None | None |
| No telemetry toggle | Yes | No | No | No |
| Image generation | Included | Included | Limited | Included |
| Code interpreter | Included | Included | Included | Included |
| Document storage | 15 GB | Limited | Limited | 2 TB (Google One) |
| Open-weight models | Yes | No | No | Partially |
Mistral Pro is the lowest-priced option among the four at $14.99, and it is the only one offering a student discount and a no-telemetry privacy toggle. The trade-off is that Mistral's ecosystem is younger and its third-party integrations are less mature than those from OpenAI or Google.
Who Should Upgrade?
Limitations and Considerations
Frequently Asked Questions
Video Resources
Before You Use AI
Your Privacy
AI tools process your inputs on remote servers unless you self-host. Free tiers may use your data for model training. Enterprise tiers typically offer data isolation and zero-retention options. Review each vendor's data processing agreement before submitting sensitive information. Mistral's Pro plan includes a No Telemetry toggle; Team accounts are opted out of training by default.
Mental Health & AI Dependency
AI assistants are tools, not therapists or companions. If you or someone you know is in crisis, contact the 988 Suicide & Crisis Lifeline (call or text 988), SAMHSA (1-800-662-4357), or Crisis Text Line (text 741741). See the NIST AI Risk Management Framework for organizational AI governance guidance.
Your Rights & Our Transparency
Under GDPR and CCPA, you have the right to access, correct, and delete your data from AI services. This article is editorially independent and not sponsored by Mistral AI. Some links may be affiliate links. See the EU AI Act for regulatory context on AI transparency obligations.