MistralLarge4.org is an independent website where you can chat with Mistral Large 4 in your browser and read guides about the model: what it costs, how it scores, how to call its API from code and what hardware it needs on your own GPUs. The site is run by suifeng.
What you can do here
- Chat with Mistral Large 4 on the homepage. A free account gets 15 credits every day, and the Pro plan covers longer conversations.
- Estimate API costs with the calculator on the API pricing page, using the prices Mistral AI publishes.
- Compare scores and specs on the benchmarks, DeepSeek V4 and Mistral Large 3 pages.
- Plan a deployment with the API guide and the run locally hardware guide.
How the chat works
Messages you send are passed to Mistral AI's official API, using the model name mistral-large-4, and the reply streams back to your browser. Each reply is billed from the tokens it actually used, and your chat history is kept in your own browser rather than on our servers. The privacy policy explains this in full.
Where our numbers come from
Every price, specification and score on the site comes from a public source, and each table names its source and the date we checked it:
- Mistral AI's launch announcement, documentation and pricing page.
- The OpenRouter and Vercel AI Gateway model listings.
- The Mistral Large 4 page on Hugging Face.
- Independent results from Artificial Analysis and Vals AI.
When Mistral AI changes a price, publishes the open weights or releases new results, we update the affected pages and change the date shown at the top of each one. Figures we calculate ourselves, such as weight memory at different precisions, show the formula next to the result.
Independence
MistralLarge4.org is not affiliated with, endorsed by or sponsored by Mistral AI. Mistral and Mistral Large are trademarks of Mistral AI and are used here only to describe the model.
Contact
Questions, corrections or partnership requests: support@mistrallarge4.org.