DeepSeek V4.1 Flash takes first on Artificial Analysis' new v4.3 benchmark, dethroning Astra
Randomdotmath · reddit · 2026-09-14
Artificial Analysis shipped a brand-new private eval in its Intelligence Index v4.3 update, replacing τ³. Astra had been farming points on it to close the gap with Fable, and the index was changed twice in three days — but the latest results show DeepSeek V4.1 Flash quietly taking first place, reshuffling the leaderboard yet again.
More from Models
- Your AI agent would cost 260% more without prompt caching, Arena data shows — arena · 2026-10-30
- EachLabs launches Lockstep recast model claiming 75% cost cut and better quality — miilesus · 2026-10-30
- ChatGPT cited a 300+ page government legal PDF it retrieved on its own — lilyraynyc · 2026-10-30
- Best MLX model for French-to-English translation on an M2 Mac? — greenreddits · 2026-10-30
- AI-driven cyber ops near the sophistication peak set by GTG-1002's Claude Code campaign — Jsevillamol · 2026-10-30
- Cognition FrontierCode chart pits Mistral Large 4 vs Chinese open models, sparking debate — onetwoval · 2026-10-30