Muse Wonder
Announced on meta.ai
Benchmark results
| Benchmark | Score | Source check |
|---|---|---|
| OSWorld 2.1Full set, partial score · max effort; mean per-task partial reward; 108 tasks | 72.6% | CorroboratedThe same number is printed in 1 other lab document: 1 |
| OSWorld 2.1Full set, strict pass rate · max effort; strict binary completion; 108 tasks | 34.3% | CorroboratedThe same number is printed in 1 other lab document: 1 |
| DeepSWE v1.1run by Datacurve; max effort | 78.8% | CorroboratedThe same number is printed in 1 other lab document: 1 |
Sources
Reported by Meta
- Muse Wonder: A Frontier Model For Personal & Professional AI Agentsresearch.meta.ai · 12 Oct 2026
Original image
