Claude Opus 5
LLM ModelsAnthropic's July 2026 flagship model, narrowly leading Claude Fable 5 on the Artificial Analysis Intelligence Index while targeting advanced coding and knowledge work.
Anthropic's flagship research partner: strong enough to lead a crowded leaderboard, but only by the narrow margin that makes the next benchmark matter.
Claude Opus 5 is Anthropic's July 2026 flagship large language model. In the July 24–25 snapshot of the Artificial Analysis Intelligence Index, it scored 61, one point ahead of Claude Fable 5 at 60 and two points ahead of GPT-5.6 Sol at 59. The narrow lead made Opus 5 the highest-scoring model in that snapshot.
The model is positioned for difficult software engineering, long-horizon agentic workflows, and professional knowledge work. Its benchmark profile is not a clean sweep: Fable 5 remains ahead on SWE-bench Pro in the comparison published around launch, while Opus 5 leads on several composite and coding-oriented measures. The result is best read as a close frontier snapshot rather than a permanent ranking.
Opus 5 also kept Anthropic's premium flagship positioning while matching the previous Opus rate card. That combination of a small Intelligence Index lead and lower measured task cost than Fable 5 made the launch more significant than the muted reaction suggested.
References & Resources
Last updated: August 3, 2026