LATEST: President Trump says the White House considers anyone who uses the term "Artificial Intelligence" instead of "Super Intelligence" as THE ENEMY.
JUST IN: OpenAI's AI found a slightly faster way to compute the Fourier transform than the classic FFT.
New bound: O(n (log n)^0.9999999999999), down from O(n log n).
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 Cloud infrastructure.
• Available to all via API today. Working with cybersecurity partners privately.
Open weights release end of October.
@Mr_Settle 5.6 Sol currently has results across more of the benchmarks we track. 6.1 Sol has fewer reported results, which pulls down its aggregate score and ranking for now.
This should change once our independent benchmarking of 6.1 Sol completes.
@koltregaskes Totally agree. The ultimate benchmark is how much you like using it vs the previous model you used as main!
All other numbers are just proxies for this.
@OpenAI 's Astra is now the #1 model on the LLM Stats coding index. And #1 on our general index too.
We estimate this is the best model and a step change in capabilities, specifically for autonomous work.
Excited to test soon, great work @sama@merettm@markchen90 and OAI team.