Opus 5 appears to lead a broad benchmark sweep across coding, search, and reasoning
signulll · x · 2026-07-25
Anthropic's Opus 5 is being compared against Fable 5, Opus 4.8, and GPT-5.6 Sol across a wide benchmark mix:
- Agentic terminal coding: Opus 5 leads Frontier-Bench v0.1 at 43.3%.
- Knowledge work: It scores 1861 on GDPval-AA v2, ahead of the others shown.
- Novel problem-solving: Opus 5 posts 30.2% on ARC-AGI-3, versus 1.5% for Opus 4.8 and 7.8% for GPT-5.6 Sol.
- Agentic search: Opus 5 reaches 90.8% on BrowseComp, narrowly ahead of GPT-5.6 Sol's 90.4%.
- Computer use / workflows: It is strong on OSWorld 2.0 (70.6%) and AutomationBench (26.0%).
- Coding: GPT-5.6 Sol leads DeepSWE v1.1 (72.7%), while Opus 5 is close at 68.8%.
- Health and biology: Opus 5 shows 59.8% on HealthBench Professional and 49.4% / 90.1% human-solved on BioMysteryBench hard.
The image also shows Opus 5 ahead or near the top on several business and legal tasks, reinforcing the claim that it is a broad capability jump rather than a narrow coding win.
Related event: Anthropic Releases Claude Opus 5: SOTA Performance at Half the Price(128 posts)→
More from Models
- Together AI sits down with Meta to explain Muse Spark Open — togethercompute · 2026-10-27
- GPT-6.1 Terra hits ~60% on Agents' Last Exam at $1/task, 9x cheaper than Opus 5.5 — ChrisGPT · 2026-10-27
- ChatGPT keeps telling tech-support users to email 404 Media's Jason Koebler — petrusenko_max · 2026-10-27
- GPT-6.1 Terra called the awkward middle tier, priced at half of GPT-6.1 Sol — cedric_chee · 2026-10-27
- Mystery model "Terra" scores 48 on Artificial Analysis index with no public API — haider1 · 2026-10-27
- GPT-6.1 Terra (Max) Appears on Artificial Analysis Scoring 48 — theanonymousone · 2026-10-27