Mistral Large 4: what the license allows, and what the maker charges
thecontinuitylayer.com/open-models/mistral-large-4Mistral AI's 1-trillion-parameter natively multimodal mixture-of-experts model with about 52B active parameters and a 1M context, released on Hugging Face in BF16 and NVFP4; Mistral Small 4 (119B, Apache 2.0) is the smaller open release.
- 1Partly
Is Mistral Large 4 under an open license?
Partly, by Mistral AI's own page.
Mistral Large 4 is under the Mistral Ecosystem Open-Weight License, which calls itself "a modified version of the MIT license" and requires a separate commercial license from Mistral before operating the model as a service for third parties. Mistral Small 4 is under Apache 2.0.
Yes means: The weights are under an OSI-approved license such as Apache-2.0 or MIT. Partly means the maker's own license with conditions, such as a user cap or an acceptable-use clause; no means research or non-commercial use only.
- 2Yes
Can you download Mistral Large 4's weights?
Yes, by Mistral AI's own page.
BF16 and NVFP4 weights are on Hugging Face under mistralai.
Yes means: The maker's page links the weights for download with nothing more than a click-through license. Partly means an access form, an approval step, or only some sizes released.
- 3Yes
Does the maker state Mistral Large 4's context window?
Yes, by Mistral AI's own page.
The card states "1M context length."
Yes means: A token count for the context window is on the maker's page or model card.
- 4Partly
Does the maker say what it takes to run Mistral Large 4 on your own machine?
Partly, by Mistral AI's own page.
About 1T total and 52B active parameters; the NVFP4 checkpoint targets Blackwell GPUs with a smaller memory footprint, the serving example uses tensor-parallel size 8, and the card gives no GPU memory requirement.
Yes means: The page states the parameter count and the memory or hardware it needs. Partly means the parameter count alone.
- 5Yes
Does the maker publish a price for its hosted Mistral Large 4?
Yes, by Mistral AI's own page.
Mistral's API pricing page lists Mistral Large 4 at $0.7 per million input tokens and $2.1 per million output tokens, cached input $0.07; Mistral Small 4 at $0.15 in and $0.6 out.
Yes means: A per-token price on the maker's own pricing page. Partly means a hosted version exists but the price sits behind sign-in or a sales call.
- 6Partly
Does the maker train on your prompts when you use its hosted Mistral Large 4?
Partly, by Mistral AI's own page.
Mistral's help center says opting out of training for la Plateforme and the APIs is a toggle in the Admin Console's Privacy panel, that Teams and Enterprise are opted out by default, and that Free, Pro and Student plans must opt out themselves.
Yes means: Here no is the good answer: the maker's page says what you send its API is not used for training. Yes means it is unless you opt out; partly means it depends on the plan or the page does not give the default.
- 7Partly
Can you fine-tune Mistral Large 4 and keep the result?
Partly, by Mistral AI's own page.
The license grants the right to "further train or otherwise modify the Model", create derivative works and "distribute, sublicense, and/or sell copies of the Model" with the notice kept; offering the model or its derivatives as a service to third parties needs a separate commercial license.
Yes means: The license allows fine-tuning and the fine-tuned weights are yours to keep and share. Partly means conditions follow the derivative, or a hosted fine-tune you cannot download.
License as stated: Mistral Ecosystem Open-Weight License, "a modified version of the MIT license"; Mistral Small 4 is Apache 2.0.. Every grade above is read from Mistral AI's own page on the day shown. If a page has changed, tell us.