deepseek-v4.1IMAGE
DeepSeek V4.1 Flash 1017 coding benchmarks beat Opus 5
Lists three head-to-head scores against Claude Opus 5: real bug fixes 77.0 vs 74.0, task automation 60.5 vs 50.3, long agentic workflows 35.4 vs 28.6 — and notes this is only the Flash tier.
MUBARANKY
@mubaranky411Archived 10 days agoOriginally posted 2026.10.17
MUBARANKY @mubaranky411Originally posted 2026.10.17
DeepSeek just dropped V4.1 Flash 1017 and the coding benchmarks are wild 😳 It's outperforming Claude Opus 5 across key tasks: • Real bug fixes: 77.0 vs 74.0 • Task automation: 60.5 vs 50.3 • Long agentic workflows: 35.4 vs 28.6 …and this is literally just their Flash model
x.com/mubaranky411/status/2111460777395626013Archived 2026.10.18 16:36