Mistral AI API Pricing
European-built AI models — 7 established models plus the Large 4 GA release
Mistral Large 4: GA API and released weights
Checked November 7, 2026: the official model card labels mistral-large-4 GA, with 1M context and text/image input. Input/cached-input/output rates are $0.70/$0.07/$2.10 per 1M tokens. These are the current displayed direct rates, not the earlier two-week preview offer.
Official weights were released October 29. The card explains 49B active parameters per token, or 52B including embeddings and output layers. Its Mistral Ecosystem Open-Weight License requires separate commercial permission for Model-as-a-Service businesses. No independently tested GPU configuration is claimed. See the maintained model guide.
| Model | Input (per 1M tokens) | Output (per 1M tokens) | Context Window | Tier |
|---|---|---|---|---|
| General Purpose | ||||
| Mistral Large 4 (GA) | $0.70 | $2.10 | 1M | GA; weights available |
| Mistral Medium 3.5 | $1.50 | $7.50 | 128K | Mid |
| Mistral Large 3 | $0.50 | $1.50 | 262K | Budget |
| Mistral Small 4 | $0.15 | $0.60 | 128K | Budget |
| Reasoning (Magistral) | ||||
| $2.00 | $5.00 | 256K | Mid | |
| $0.50 | $1.50 | 256K | Budget | |
| Coding (Devstral) | ||||
| $0.40 | $2.00 | 128K | Budget | |
| $0.10 | $0.30 | 128K | Budget | |
| Codestral | $0.30 | $0.90 | 256K | Mid |
| Edge (Ministral) | ||||
| Ministral 3 14B | $0.20 | $0.20 | 128K | Budget |
| Ministral 3 8B | $0.15 | $0.15 | 128K | Budget |
| Ministral 3 3B | $0.10 | $0.10 | 128K | Budget |
Why Mistral?
European-built AI models designed for cost-conscious developers who need reliable performance.
European Data Sovereignty
Mistral is headquartered in Paris and stores data in Europe. Ideal for GDPR-sensitive workloads and EU-based companies that need to comply with strict data residency requirements.
Open-Source Heritage
Mistral releases open-weight models alongside their API, giving you the flexibility to self-host or switch providers. No vendor lock-in — deploy on your own infrastructure if needed.
Competitive Pricing
Established options include Large 3 at $0.50/$1.50 and Ministral 3 3B at $0.10/$0.10 per 1M tokens. Large 4 adds a GA flagship option at $0.70/$2.10 (current flat first-party rate, checked Oct 30, 2026 — it replaced the Oct 6–20 launch offer and the $1.36/$4.18 list rates seen Oct 21). Retired Devstral and Magistral API rates are historical, not currently purchasable endpoints.
Which Mistral Model Should You Use?
Mistral offers four model families for different workloads.
🧠 Magistral (Reasoning)
- 256K context window
- Specialized for complex reasoning
- Math, coding, and logic tasks
- Open-weight under Apache 2.0
💻 Devstral (Coding)
- 128K–256K context window
- Code generation and software engineering
- Devstral Small 2 is cheapest coding model
- Codestral for code-specific tasks
Established options include Large 3 at $0.50/$1.50 and Ministral 3 3B at $0.10/$0.10 per 1M tokens. Large 4 adds a GA flagship option at $0.70/$2.10 (current flat first-party rate, checked Oct 30, 2026 — it replaced the Oct 6–20 launch offer and the $1.36/$4.18 list rates seen Oct 21). Retired Devstral and Magistral API rates are historical, not currently purchasable endpoints.
⚡ Mistral General Purpose
- 128K–1M context, depending on model
- Large 4 GA, Large 3, Medium 3.5, Small 4
- Multilingual and multimodal
- Open-weight models available
📱 Ministral (Edge)
- 128K context window
- 3B, 8B, and 14B parameter sizes
- Optimized for on-device deployment
- Low latency and cost
Mistral Cost Calculator
Estimate your monthly spend on Mistral AI models.
Monthly Cost Estimate
Based on published API pricing. Actual costs may vary.
Need the full calculator with all providers? Compare Mistral against OpenAI, Anthropic, Google, and more.
Try the Full APIpulse CalculatorMistral vs Competitors
How Mistral pricing stacks up against other major LLM providers.
| Provider | Model | Input (per 1M tokens) | Output (per 1M tokens) | Context Window |
|---|---|---|---|---|
| Mistral | Large | $0.50 | $1.50 | 128K |
| OpenAI | GPT-4o | $2.50 | $10.00 | 128K |
| Anthropic | Claude Sonnet 4.6 | $3.00 | $15.00 | 200K |
| Gemini 2.5 Pro | $1.25 | $10.00 | 1M | |
| Mistral | Small | $0.15 | $0.60 | 128K |
| OpenAI | GPT-4o mini | $0.15 | $0.60 | 128K |
| Gemini 2.5 Flash-Lite | $0.10 | $0.40 | 1M | |
| Anthropic | Claude Haiku 4.5 | $0.80 | $4.00 | 200K |
See how Mistral models stack up against OpenAI, Anthropic, and more
Related Reading
Deep dives into Mistral pricing and how it compares to other providers.
Mistral vs GPT-4o: Can the European Challenger Beat OpenAI on Price?
A pricing and capability comparison between Mistral Large 3/Small and GPT-4o/GPT-4o mini for cost-conscious developers. Includes real-world cost breakdowns for chatbots, code generation, and document analysis.
Read the full comparison →Calculate Your Mistral Costs
See exactly what Mistral will cost for your specific workload. Compare against other providers in seconds.
Go to Calculator →