Free
Mistral Large 4 is free to call on NVIDIA, no per-token billing
Mistral's new flagship, Mistral Large 4, now has a free endpoint on NVIDIA. Sign up for an NVIDIA account, generate an API key and call it on build.nvidia.com for free at about 40 requests per minute, great for prototyping and testing.
Free access to Mistral Large 4 at about 40 RPM, no per-token billing
- Eligibility：
- Create and verify an NVIDIA account and join the free NVIDIA Developer Program. Free endpoints are for prototyping, research and testing; production use serving real users requires an NVIDIA AI Enterprise license.
- Benefit duration：
- Available while you are a Developer Program member; rate limits are set per model and account
The briefing
Mistral's new flagship model, Mistral Large 4, is now in NVIDIA's API catalog with a free endpoint. Sign up, generate an API key and start calling it. No card required and no per-token billing.
What you can use
Mistral Large 4 is a mixture-of-experts model with about 1 trillion total and 52 billion active parameters. It takes text and image input, supports reasoning, function calling and structured outputs, and has a context window of about 262K tokens.
The free endpoint is rate limited to about 40 requests per minute; your own limit is shown in the top-right corner of the dashboard. Free endpoints are for prototyping, research and testing. Production use serving real users needs an NVIDIA AI Enterprise license.
How to use it
- Open the Mistral Large 4 model page, sign in with your NVIDIA account and try it in the Playground.
- Click Generate API Key, then call the OpenAI-compatible endpoint
https://integrate.api.nvidia.com/v1/chat/completionswith the model namemistralai/mistral-large-4-1t-a52b.
For more free models, see our NVIDIA API Catalog free models roundup. Full terms are in the NVIDIA NIM FAQ.