Mistral Large 4: price, features and how it is used
Mistral Large 4 is a natively multimodal MoE model with about 1 trillion parameters, of which 52 billion are active. It was announced on 6 October 2026 as a public preview, and the documentation now lists it as generally available. The license is the Mistral Ecosystem Open Weight License v26.10.
Provider: Mistral AI · Released: October 2026 · open weights
Results by test
| Test | Result | Source |
|---|---|---|
| DeepSWE 1.1 | 61,7% | Mistral, announcement of Mistral Large 4 |
| SWE-Atlas-QnA | 59,4% | Mistral, announcement of Mistral Large 4 |
| Terminal-Bench 4 | 28,3% | Mistral, announcement of Mistral Large 4 |
| AutomationBench | 59,9% | Mistral, announcement of Mistral Large 4 |
| Cybench | 93% | Mistral, announcement of Mistral Large 4 |
Intelligence, price and speed by effort level
There is no data yet for this model on effort level, intelligence and price per task.
Data on the Intelligence Index, capability indexes, price per task, speed and latency: Artificial Analysis. Prices per 1M tokens are from official vendor data.
Compare Mistral Large 4 with another model
Choose a model and see prices, context and results side by side.
How to use Mistral Large 4
Mistral Large 4 is available via public preview on the API in Mistral Studio (console.mistral.ai). Le Chat mobile apps are mentioned in the announcement, and Mistral says there will also be a European deployment run by the company itself.
- Create an account in Mistral Studio and an API key.
- Call /v1/chat/completions or /v1/conversations with the model mistral-large-4.
- If you want a ready-made chat, the announcement points to the Le Chat mobile apps. Plans and limits are not described there.
- The license is “GA Mistral Ecosystem Open Weight License v26.10”. The announcement says the weights will be released by the end of the month, so check whether they are already available if you want to self-host.
- For non-English text, keep in mind that non-English text uses more tokens than English, so the same text costs more. Check the quality of answers on your own real examples.
What it is suited for
- Low price: $0.70 input and $2.10 output per 1M tokens, cached input $0,07.
- Context window of 1M tokens.
- Open weights under the Mistral license.
What to keep in mind
- The license is not a standard open license but a specific Mistral license.
- In the announcement the model is in public preview. The documentation dated November 5, 2026 lists it as GA.
- Compared with GLM-5.3, Mistral reported a result on par with or close to it in finance and coding, i.e. with no clear advantage.
- The results are from the provider itself.
Frequently asked questions about Mistral Large 4
- How much does Mistral Large 4 cost?
- 0.70 dollars per 1M input tokens, 0.07 dollars for cached input and 2.10 dollars per 1M output. The context is 1M tokens.
- Is Mistral Large 4 free?
- Free access is not confirmed, because the announcement does not describe Le Chat plans. The API is billed per token and the weights have an open license, but hosting is your own cost.
- Does Mistral Large 4 support multiple European languages?
- Yes, according to Mistral, training covered over 160 languages, including all official EU languages. Still, check the quality on your own texts, because tokens in many languages are more numerous than in English.
- Can Mistral Large 4 be run on your own server?
- The model is open-weight under the GA Mistral Ecosystem Open Weight License v26.10. However, the announcement says the weights are promised by the end of the month, so check first whether they have been published.
More models in AI Benchmark
- Gemini 4 Neon (Google)
- Gemini 4 Carbon (Google)
- GPT-6 Astra (OpenAI)
- GPT-6.1 Sol (OpenAI)
- GPT-6.1 Terra (OpenAI)
- Muse Wonder Full (Meta)
Intelligence Index, token, TTFT and effort are explained in the AI Benchmark glossary. Prices are per 1M tokens and may change, so check the provider’s official page.
Sources: Mistral, Large 4 announcement · Mistral, model documentation · Mistral, changelog
Last checked: 18.11.2026
Compare with other models
Set Mistral Large 4: price, features and how it is used to another model by price, context and benchmarks.