Arthur Mensch told TechCrunch there’s “a very exciting model to come this summer — it will be open-weight, and we’re opening early access to it in July.” No specs, no benchmarks, no name. But the signal is loud: Europe’s biggest OpenAI challenger is betting the open camp can still catch the closed frontier.
What Mistral is actually shipping
A large language model you can download and run yourself — not a closed API you rent. That’s the entire pitch. Mistral already ships Large 3, a 675B-parameter mixture-of-experts model (41B active) under Apache 2.0, scoring ~73% on MMLU-Pro. The summer model is the step up, teased on X under the “Le Chaton Fat” nickname and going first to key partners in early access.
Why it’s worth watching
Follow the money: ARR jumped from $20M to over $400M in a year, on track to pass $1B, with a reported $3.5B raise on the table. Open weights plus enterprise cash is a rare combo. If closed labs win by renting you intelligence, Mistral’s whole thesis is that you shouldn’t have to. Expect access via the La Plateforme API and self-hosting once it lands — July’s early access is the real test of whether open can keep pace.
You Might Also Like
- Openai Trusted Access for Cyber Opens gpt 5 5 to Offensive Security Work for Verified Defenders Only
- Openai gpt 5 6 Government Managed Access List 20 Partners Approved one by one
- Gpt oss 120b Openai Finally Goes Open Source and its Worth the Wait
- Openai Symphony Finally a Framework That Lets you Stop Babysitting Your Coding Agents
- Openai Just Acquired Promptfoo the 86m ai Security Startup Used by 25 of Fortune 500
Top AI Product 
Leave a comment