Mistral's free plan includes $10 a month in API credits, enough to try the new Mistral Large 4

Mistral released a public preview of Mistral Large 4 on October 6. It's Mistral's largest model so far, with about 1.05 trillion total parameters, 49 billion active parameters, native image input, and a 1M-token context window. The weights are expected to be released at the end of the month, and you can already try it through the API in Mistral Studio.

Mistral's free plan includes $10 in API credits every month, which is a handy way to get started with the new model.

What you get

New Mistral accounts start on the free plan by default, with no subscription needed. The free plan comes with $10 a month in API credits, shared across Studio, the API, and Vibe Code. Once the credits run out, usage stops until the next billing period unless you turn on pay-as-you-go.

Free-plan rate limits are the lowest tier and are meant for testing and prototyping. If you send requests too quickly you'll get a 429 error, so just wait a moment and retry. The free plan also includes limited chats, web searches, and coding sessions in Vibe on web and mobile.

What you can use

The model ID for Mistral Large 4 is mistral-large-4. Input costs $1.36 per million tokens, cached input $0.14 per million, and output $4.18 per million, with batch processing at half price. At those rates, $10 covers roughly 2.4 million output tokens.

It supports function calling, structured outputs, and document Q&A, and Mistral's launch focuses on its results in coding, agentic tasks, and cybersecurity. Mistral Medium, Mistral Small, OCR, and the Voxtral speech models in the same account draw from the same credits.

How to claim

1. Sign up in Mistral Studio and activate Studio on the free plan.
2. Create a key on the API Keys page and set the model to mistral-large-4. If you'd rather not write code, you can chat with it in the Studio Playground first.

Details: Mistral Large 4 announcement, model and pricing, and Mistral plans and pricing.