Leaked benchmark chart shows Meta's "Muse Wonder" topping agent evals vs Claude and GPT
zainhas · x · 2026-10-13
A comparison chart shows a model codenamed "Muse Wonder Full" from Meta leading several benchmarks: Personal Superintelligence Index 59.1 (vs Claude Opus 5.5's 55.2 and GPT-6 Astra's 54.5), plus top scores on DeepSWE (79.0), SWE-Marathon (57.5), and AutomationBench (45.4); Harvey LAB-AA 13.1 far ahead of rivals. It still trails on Terminal-Bench 4.0, HealthBench Pro, ProgramBench, and the Artificial Analysis Index. Note: chart of unclear provenance, likely a leak — unverified.
More from Models
- Google AI Overviews denied a person had died, renewing criticism of AIO for news — lilyraynyc · 2026-10-13
- User feeds Claude Opus GTA frame-by-frame screenshots since it can't watch videos — adonis_singh · 2026-10-13
- Gemini Deep Research on web is really slow today, user reports — gaganghotra_ · 2026-10-13
- Dual RAID10 NVFP4 setup loads GLM 5.3 Flash cold in ~1 minute — TheZachMueller · 2026-10-13
- Opus 5.5 stuns at SVG generation; Meta's Muse Wonder judged 3-6 months behind frontier — scaling01 · 2026-10-13
- Muse vs Opus 5.5: same 3D prompt costs $4.11 vs $99.43, big quality gap — iamaliveix · 2026-10-13