deepseek-v4.1IMAGE
DeepSeek V4.1 Flash edges Opus 5.5 on a 105-bug repair test at $1.84
On 105 reproduced bugs across two repos, DeepSeek V4.1 Flash 1017 scored 42.7, slightly above Opus 5.5's 41.7 and Kimi K3.1 Pro's 36, while costing $1.84 versus Opus's $58.53.
Leo · Building w/ AI
@leo114119Archived 8 days agoOriginally posted 2026.10.24

Leo · Building w/ AI @leo114119Originally posted 2026.10.24
Kimi K3.1 Pro 真实任务测试 (2 个仓库,105 个重现的 bug) 看看大模型能修复几个,同时看看要花多少钱 结果如下: DeepSeek V4.1 Flash 1017 (max):42.7 Muse Wonder Full 1.0 (max):42 Opus 5.5 (max):41.7 Kimi K3.1 Pro (max):36 Muse Spark 1.3 (max):32.2 成本如下: DeepSeek V4.1 Flash 1017 (max):$1.84 Opus 5.5 (max):$58.53 Kimi K3.1 Pro (max):$13.76 感觉在真实的编码任务里,也不怎么样 还是那句话,对比K3有进步,但完全比不过DeepSeek V4.1 Flash 1017 https://x.com/PawelHuryn/status/2113747367409123623/photo/1
x.com/leo114119/status/2113796740713246914Archived 2026.10.24 12:05