gpt-6POST
AA leaderboard: Kimi K3.1 Pro at 55 edges GPT-6 Astra's 53
The author cites AA leaderboard scores to rebut claims of an open-source model crisis: Opus 5.5 at 58, Kimi K3.1 Pro at 55, and GPT-6 Astra at 53, putting a Chinese open model fourth globally.
Patrick 老派
@laopai1990Archived 4 days agoOriginally posted 2026.10.24
Patrick 老派 @laopai1990Originally posted 2026.10.24
有人说，Kimi K3.1 Pro 发布后，开源模型反而陷入危机了。 我看了一眼 AA 榜单： Opus 5.5：58 Kimi K3.1 Pro：55 GPT-6 Astra：53 不是哥们，国产模型都冲到全球第四了，怎么还叫危机？？ 难道现在开源模型不拿第一，都算失败了吗？🤣
x.com/laopai1990/status/2113844585718022437Archived 2026.10.25 10:05