Mistral AI launches Mistral Large 4, a 1-trillion-parameter model nicknamed Le Chonk
Decrypt reported that Paris-based Mistral AI launched Mistral Large 4 on October 6, a 1-trillion-parameter model that activates 49 billion parameters per query and is nicknamed Le Chonk.
Why it matters
Nothing Decrypt reported connects the model to crypto markets, token prices or trading, so the relevance to this desk is limited. The open-weight framing and the promised weights release by the end of October are the parts with a date attached, and both come from Mistral itself.
Key facts
Mistral AI launched Mistral Large 4 on October 6 as a model with 1 trillion parameters.
1,000,000,000,000 parametersReported by Decrypt
Large 4 routes each query to specialist sub-networks and fires 49 billion of its parameters per answer.
49,000,000,000 parametersReported by Decrypt
Mistral charges $1.36 per million input tokens and $4.18 per million output tokens for Large 4.
1.36 USDReported by Decrypt
On the Finance Agent v2 chart, Large 4 scores 54.7, above GPT-6 Astra at 53.5 and below Claude Opus 5.5 at 58.6.
54.7Reported by Decrypt
Mistral says it will release the Large 4 weights by the end of October.
Reported by Decrypt
Mistral AI chief scientist Guillaume Lample wrote on X that Large 4 is the strongest open-weight model from the US or Europe.
Reported by Decrypt
The previous flagship Large 3 used the same design with 41 billion active parameters out of 675 billion.
675,000,000,000 parametersReported by Decrypt
What happened
Paris-based Mistral AI launched Mistral Large 4 on October 6, and Decrypt reported it as a model with 1 trillion parameters that activates 49 billion of them per answer through a mixture-of-experts design that routes each question to a few specialist sub-networks. Decrypt put the price at $1.36 per million input tokens and $4.18 per million output tokens. Mistral AI chief scientist Guillaume Lample wrote on X that the model is at the frontier of open models and is by far the strongest open-weight model from the US or Europe, which is the company's own claim rather than an independent assessment. The name Le Chonk comes from Le Chaton Fat, a fictional model with 30 trillion parameters that Reddit and X users invented in June after Mistral renamed its Le Chat assistant to Vibe. On the Finance Agent v2 launch chart cited by Decrypt, Large 4 scores 54.7, ahead of GPT-6 Astra at 53.5 and behind Claude Opus 5.5 at 58.6. Mistral says it will release the model's weights by the end of October. Decrypt is the only outlet in this set covering the launch, so the pricing, the benchmark scores and the strength claim all rest on a single report.
How the story developed
- First report
First report by Decrypt
Mistral AI Drops 'Le Chonk': A Massive AI Model Named After a Cat Meme
- Update
Mistral AI launches Mistral Large 4
Decrypt reported a 1-trillion-parameter model with 49 billion active parameters per query, priced at $1.36 per million input tokens and $4.18 per million output tokens.
- Official statement
Guillaume Lample posts the open-weight strength claim on X
The chief scientist called Large 4 the strongest open-weight model from the US or Europe; the statement is Mistral's own.
Affected entities
Original sources
| Publisher | Report | Role | Published |
|---|---|---|---|
| Decrypt Crypto media | Mistral AI Drops 'Le Chonk': A Massive AI Model Named After a Cat Meme | Original |