Claude Opus 5 Hits 70.6% on OSWorld v2, Dev Offers Bounty for Harder Evals
EricBuess · x · 2026-07-25
Claude Opus 5 has achieved a state-of-the-art 70.6% score on the recently released OSWorld v2 computer use eval, a significant increase from 55.7%. The developer notes that current benchmarks are becoming too easy and is offering to sponsor a harder computer use eval where Opus 5 scores below 50%.
More from Models
- Opus 5 reportedly edges past downgraded Fable 5 at half the price — oran_ge · 2026-07-25
- ‘The only eval I pay attention to’: Opus 5 and a RunescapeBench joke — Flomerboy · 2026-07-25
- GPT 5.6 Becomes More Casual in Chat, Opus 5 Can Feel Testy — dreamwieber · 2026-07-25
- Miles Brundage says Opus 5 copies joke setups, then offers to “be original” — Miles_Brundage · 2026-07-25
- Kimi K3.0 abliterated GGUF starts trending on Hugging Face — audnai · 2026-07-25
- Early reaction to Opus 5: colder, less pushback than the 4.7–4.8 line — teortaxesTex · 2026-07-25