Claude Opus 5 tops several benchmarks, including 43.3% on agentic terminal coding
Byakko_4 · reddit · 2026-07-25
- A benchmark graphic for Claude Opus 5 compares it with Fable 5, Opus 4.8, and GPT-5.6 Sol.
- It shows 43.3% on agentic terminal coding, 1861 on knowledge work, 30.2% on ARC-AGI-3, 90.8% on agentic search, and 70.6% on computer use.
- The image also suggests a mixed picture: Opus 5 leads on several categories, while GPT-5.6 Sol is ahead on DeepSWE v1.1 and Fable 5 slightly edges Opus 5 on Humanity’s Last Exam without tools.
Related event: Anthropic Releases Claude Opus 5: SOTA Performance at Half the Price(128 posts)→
More from Models
- Same Echo Maze prompt, three frontier models: all passed visually but shipped the same hidden bug — eyishazyer · 2026-09-11
- Benchmark scores drop from 89% to 19% on new evals — how benchmaxxing breaks leaderboard trust — airesearch12 · 2026-09-11
- ChatGPT tells user their question is too hard and to 'accept dumber answers' — phido3000 · 2026-09-11
- Claude is no longer available for minors as Anthropic rolls out age assurance — Muhammad523 · 2026-09-11
- Developer Building a Unified Leaderboard of All Model Benchmark Scores — airesearch12 · 2026-09-11
- Rumor claims Kimi faked performance by serving Claude; DeepSeek new model surprises in evals — realsohamparekh · 2026-09-11