deepseek-v4.1POST
DeepSeek V4.1-Flash-1017 goes live on web, app and API with shorter outputs
DeepSeek says the new version scores higher, finishes tasks faster and answers more briefly than the last one, calling it a Pareto improvement in performance and cost; the poster notes output length drives agent-run spend.
Kapish | AI, Actually
@kapishagarwal_Archived 12 days agoOriginally posted 2026.10.19
Kapish | AI, Actually @kapishagarwal_Originally posted 2026.10.19
deepseek pushed v4.1-flash-1017, live on web, app and api. deepseek says: better benchmark scores, finishes tasks faster, shorter answers than the last version. they're calling it a "pareto improvement in performance and cost". shorter outputs is the part to watch. on agent runs you pay for every token it rambles
x.com/kapishagarwal_/status/2112015977093156896Archived 2026.10.19 12:03