deepseek-v4.1THREAD
105 个 bug 修复实测：V4.1 Flash 42.7 分超过 Opus 5.5
在两个真实仓库的 105 个植入 bug 上，DeepSeek V4.1 Flash 1017 得 42.7，高于 Muse Wonder 42、Opus 5.5 41.7、Kimi K3.1 Pro 36。
Paweł Huryn
@PawelHuryn4 天前收录原推发布于 2026.10.24
THREAD
So I tested Kimi K3.1 Pro on real work. 2 repos. 105 replanted bugs. Find and fix what you can. The results: 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 I'm sorry. Considering the cost, I see no reason to use this model.
@dylanvdmerwe I agree. Yet, its max is worse than DeepSeek V4.1 Flash 1017 high effort. Normally, I would continue. Here, I stopped wasting tokens.
Paweł Huryn @PawelHuryn原推发布于 2026.10.24
x.com/pawelhuryn/status/2113747367409123623采集于 2026.10.24 12:07