Peaked at #1 Off the board
Mistral to Release Large 4 Open Weights
Mistral co-founder Mensch says ML4 was trained and served on its own compute, with RL showing no sign of saturation.
Key points
- Mistral AI introduced Mistral Large 4, also called Le Chonk, with 1T parameters, natively multimodal and 49B active.
- Mistral AI said it is the best open weights model from the US or Europe on aggregated benchmarks, state-of-the-art on cyber defense, manufacturing and finance, and surpasses closed frontier models on visual grounding.
- Arthur Mensch said the model was trained and served on Mistral's own compute, and that RL shows no sign of saturation.
- Julien Chaumond said the open weights release is set for end of October.
Key points and the reaction summary are written by AI from the posts on this page. Check the original post. How we use AI
Original post
That one took some groundwork. Trained and served on our own compute, and RL shows no sign of saturationQuoting @MistralAI: Meet Mistral Large 4, aka Le Chonk.
• 1T parameters, natively multimodal. 49B active.
It is the best open weights model from US or Europe on aggregated benchmarks.
• State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding.
• Forged in Europe end-to-end and is deployable from Europe via our own Mistral
What others are saying
2 more-
Julien Chaumond quoted Mistral Large 4 (1T params, 49B active, natively multimodal) announcing open weights release at end of October.
View on X -
Mistral researcher says they released Large 4 preview, the only Western open-source LLM competitive with top Chinese open-source models on coding.
View on X
Top replies on X
Sign in to see 5 top replies from X
From @petrroyce, @LexnLin, @KimNoel and others. Spam removed, with English and Chinese translations.
Discussion 0
Sign up Sign in to join the discussion
No comments yet. Start the conversation.