Strengths
- Open weights under a modified MIT license. Mistral says four GPUs are enough to run it yourself.
- One model for everyday tasks, reasoning and coding. You turn thinking on or off with each request.
- Mistral reports 77.6% on the SWE-bench Verified coding test.
- It is the default model in Vibe, so you get it on the free plan too.
- A 256K token context window, and it understands both text and images.
Weaknesses
- Much pricier than Chinese open models: $1.50 input and $7.50 output per million tokens. The far smaller Qwen 3.6, with 27 billion parameters, scores 72.4% on SWE-bench Verified.
- Output costs more than on the newer Mistral Large 4: $7.50 versus $4.18 per million tokens.
- The license is not fully permissive: companies with large revenue fall outside the free terms.
Best for
- an assistant and coding agent with data kept in the EU
- self-hosting on 4 to 8 GPUs
- company work with sensitive data where GDPR matters
Not ideal for
- the cheapest bulk processing
- large companies that want a permissive license without negotiating
How to use it in Europe
The easiest way to try it is the Vibe app on the web or mobile, where it is the default model even on the free plan. The mobile app is still called Le Chat in the app stores. Developers use it through the Mistral API, and the weights are on Hugging Face. Mistral stores data in the EU by default. On Free and Pro, conversations may be used for training until you turn this off in the privacy settings.
Price compared with similar models
Benchmark results
| Benchmark | Score | Source |
|---|---|---|
| SWE-bench Verified | 77.6% | huggingface.co ↗ (vendor-reported), April 2026 |
| tau3-Telecom | 91.4% | huggingface.co ↗ (vendor-reported), April 2026 |
Benchmarks cover only a narrow slice of ability and vendors often report them at their best settings. Treat them as a rough guide.
More models from this vendor
- Mistral Large 4: Mistral's largest model, still in preview. Built for demanding tasks, long documents and security analysis with data kept in the EU.
- Mistral Small 4: Mistral's cheap, fast open model under Apache 2.0 for everyday tasks through the API or on your own server.
- Ministral 3 14B: Mistral's small open model under Apache 2.0 that runs on a single GPU, even without an internet connection.
- Codestral 25.08: Mistral's model for code completion right in your editor. Fast and cheap, but dated for autonomous coding agents.
- Mistral Large 3: Mistral's previous large model with open weights under Apache 2.0. Cheap in the API and suited to your own powerful server.
Similar models from other vendors
- Claude Sonnet 5.5 (Anthropic): Claude's balanced model for everyday work and coding. It is faster than Opus 5.5 and costs half as much.
- Command A+ (Cohere): An open business model from Canada's Cohere that officially supports Slovak and Czech and runs on your own server.
- Gemini 3.8 Flash (Google): Google's fast, lower-cost model for everyday work, coding and document analysis. You can try it in the API for free.
- Muse Spark 1.3 (Meta): Meta's balanced model with a 1-million-token window and a fair price. The Muse Spark line also powers the Meta AI assistant.
- Muse Spark Open (Meta): Meta's large open model that a company can run on its own servers, so data never leaves its infrastructure.
- Llama 4 Maverick (Meta): The larger model in the newest Llama generation, from April 2025. Cheap through providers, but its license excludes people and companies in the EU.