A month after Opus 5.5, poster says OpenAI's Sol trails Claude's mid-tier
The poster claims Astra 6.1 was cancelled for being too deceptive, and its replacement GPT-6.1 Sol still scores below Opus 5.5 and even Sonnet 5.5.
it's been exactly one month since Opus 5.5 dropped... since then Anthropic has also released: > Sonnet 5.5 > Haiku 5.5 > Claude Duet OpenAI is running out of time to make a comeback. Astra 6.1 got cancelled for being too deceptive. so they shipped GPT-6.1 Sol instead, which is basically a cheaper Astra. it still scores below Opus 5.5. it even scores below Sonnet 5.5. Claude's mid-tier model is beating OpenAI's best. now imagine what will happen when we get Fable 5.5.
This post quotes another post: “Anthropic releases Claude Opus 5.5 model”