deepseek-v4.1IMAGE
Browser Use benchmark: Kimi K3.1 Pro 76.8 second, V4.1 Flash 70.6 at about $0.2 per task
On Browser Use's own benchmark Kimi K3.1 Pro ranks second at 76.8 behind GPT-6.1 Sol's 78.9 and ahead of Opus 5.5, while DeepSeek V4.1 Flash scores 70.6 as the next best Chinese model at roughly $0.2 per task.
Neo
@NeoAIForecastArchived 2 days agoOriginally posted 2026.10.27
Neo @NeoAIForecastOriginally posted 2026.10.27
Kimi K3.1 Pro is #2 on Browser Use's own benchmark. 76.8. Only GPT-6.1 Sol is higher, at 78.9. Ahead of Opus 5.5 (74.9), Grok 4.7 (73.4) and Sonnet 5.5 (72.3). About $2 a task on their chart. Next best Chinese model by their numbers is DeepSeek V4.1 Flash at 70.6, at roughly $0.2 a task
x.com/neoaiforecast/status/2114831595710615730Archived 2026.10.27 12:08