DeepSeek V4.1 Flash matches Opus 5.5 at 75/100 for 1/11 the price

Benchmark comparison: V4.1 Flash max scores 75/100 at $0.22 per task, the same as Opus 5.5 at 11x the price; only Sol and Astra score higher, at 6x+ cost.

Gregor Zunic

@gregpr07Archived 12 days agoOriginally posted 2026.10.18

Gregor Zunic @gregpr07Originally posted 2026.10.18

Open original on X

OpenAI and Anthropic should be a bit worried > new DeepSeek V4.1 Flash max: 75/100 for $0.22/task > Opus 5.5: same score, 11x the price > only Sol and Astra score higher, 6x+ the price

This post quotes another post: “DeepSeek V4.1 Flash 1017 matches Opus 5.5 on long-horizon browser use at $0.22/task”

x.com/gregpr07/status/2111697626362175787Archived 2026.10.18 16:24

Open original on X