Decision 3.0 plays games autonomously; Mistral CEO teases open SOTA model
Thom_Wolf · x · 2026-10-13
- User victormustar tested Decision 3.0 on an apple-harvesting game: given a screenshot, game stats, and memory of its last four moves, it picked its own moves and collected 7 apples in 38 moves with zero damage — no pathfinding controller needed.
- Mistral CEO ThomWolf quote-reposted it saying "open and state-of-the-art decision models - just saying", hinting at a potential open model release, unconfirmed.
More from Models
- Meta Muse retention nearly matches ChatGPT with 30%+ D7 — denny_zhou · 2026-10-13
- AA Index: Muse Wonder weak on knowledge but with very low hallucination rates — ArtificialAnlys · 2026-10-13
- Muse Wonder models now live on Artificial Analysis leaderboards — ArtificialAnlys · 2026-10-13
- Muse Wonder cuts hallucination rate to 23%, but knowledge remains a weak spot — ArtificialAnlys · 2026-10-13
- Muse Wonder Full tops CritPt and GDP.pdf, top-3 on six of ten AA evals — ArtificialAnlys · 2026-10-13
- SubQ claims 128x compute savings at 2M tokens with Subquadratic Sparse Attention — ryanallred · 2026-10-13