New todaydeepseek-v4.1POST
DeepSeek V4.1 Flash on one eval: 32 min, 220K output tokens, $0.29
On the same eval run, DeepSeek V4.1 Flash 1017 took 32 minutes and 220K output tokens for $0.29, versus 79 seconds for GPT-6.1 Terra, 16 minutes for Opus 5.5 and 4 minutes at $1.11 for Muse Wonder.
chacha
@maker_evanArchived 7 hours agoOriginally posted 2026.10.28
chacha @maker_evanOriginally posted 2026.10.28
GPT-6.1 Terra: 79 s, 3.5K output tokens (ChatGPT plan, no cost recorded) DeepSeek V4.1 Flash (1017): 32 min, 220K output tokens, $0.29 Claude Opus 5.5: 16 min, 79K output tokens (Claude plan, no cost recorded) Muse Wonder: 4 min, 17K output tokens, $1.11 All four, live: https://evals.ag/runs/run_6VWMV1M7653T9ZHE8WPCJA99N1
x.com/maker_evan/status/2115284537915969640Archived 2026.10.28 12:02