Mistral Large 4
Mistral Large 4
Mistral's new flagship: a trillion-parameter sparse model, with the weights promised by the end of October.
Updated
The read
Large 4 went into public preview on 6 October as a sparse model of just over a trillion parameters with about 52 billion active, plus a vision encoder. Mistral says the weights follow at the end of the month, and if that release is permissive it will be the strongest European open model by a distance. Today it is a hosted preview at a launch discount, trailing the leading Chinese open models on the independent indexes while keeping the European hosting story intact.
Where it shines
- Mistral's strongest model yet on independent benchmarks
- EU hosting without a bespoke contract
- A launch discount halves an already moderate price
Worth knowing first
- Still a preview, and the weights and their licence are not out yet
- Trails the leading Chinese open models on the independent indexes
Index score
82.1
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
- Rank
- 35 of 73
- Builds
- LLMs
- Category
- Frontier
- Output
- Text, code and images in, text out
- Pricing
- Paid only
- Best for
- European teams that want a frontier-scale model with EU hosting
Scores are our own editorial judgement, weighted craft 55, speed 10, control 15 and value 20.
Also worth a look
Full index- 01Claude Opus 5.5Claude Opus 5.5ProvisionalFable-level answers for less than the Opus it replaces. The new default.Frontier modelPaid only92.8
- 03Gemini 4 ArgonGemini 4 ArgonProvisionalTop of the human-preference leaderboard, and almost nobody can use it yet.Frontier modelPaid only90.6
- 04GPT-6.1 SolGPT-6.1 SolProvisionalNear-Astra answers at a fifth of the price. The OpenAI model most teams should deploy now.Frontier modelPaid only90.3
