deepseek-v4.1VIDEO
Artificial Analysis compares per-task output tokens: V4.1 Flash at 1053
Artificial Analysis shows sample submissions for one brief with cost and output tokens per task: Opus 5.5 at 1536, DeepSeek V4.1 Flash at 1053, MiniMax-M3 at 724, Nemotron 3 Ultra at 304.
Artificial Analysis
@ArtificialAnlysArchived 2 days agoOriginally posted 2026.10.27
Artificial Analysis @ArtificialAnlysOriginally posted 2026.10.27
Sample submissions for a single brief below, including the Cost and Output Tokens per Task. We include the following models: Claude Opus 5.5 (max with fallback, 1536), DeepSeek V4.1 Flash (max, 1053), MiniMax-M3 (724), Nemotron 3 Ultra (304)
x.com/artificialanlys/status/2114845078082683367Archived 2026.10.27 12:08