今天新增benchmarkIMAGE

Sonnet 5.5 在新榜单上逼近 Opus 5.5 与 Astra

用户感叹某新基准里 Sonnet 5.5 与 Opus 5.5、GPT-6 Astra 分数接近，却远超 Fable 5.1，认为 Anthropic 的 Sonnet 表现惊人。

Token Gremlin

@TokenGremlin8 小时前收录原推发布于 2026.10.28

Token Gremlin @TokenGremlin原推发布于 2026.10.28

I really like this new benchmark and have a lot of respect for the team behind it. And I have to admit, it's kinda scary seeing Sonnet 5.5 so close to Opus 5.5 and GPT-6 Astra, while being so far ahead of Fable 5.1. Anthropic really cooked with Sonnet, holy shit...

这条是引用着另一条写的：x.com/cognition/status/2115132488775524793

x.com/tokengremlin/status/2115138522629677223采集于 2026.10.28 10:11