A user compared Anthropic's Claude Opus 4.8 and Opus 5 on 25 coding tasks, finding that both models achieved a 9/25 strict test pass rate. Opus 5 demonstrated broader search and verification, using more commands and revisions, while Opus 4.8 maintained a smaller code footprint. Despite these differences in approach, the cost and performance metrics remained similar, with Opus 5 being slightly cheaper and using more tokens and time. AI
IMPACT Opus 5's broader search strategy may offer new approaches to complex coding tasks, though its practical utility remains comparable to Opus 4.8.
RANK_REASON User comparison of two model versions, not a primary release.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →