Muse Wonder (Medium): benchmark results
Provider: Meta. Released 2026-10-12. Access: API.
Aggregate Rating 1872 ± 24, rank #64 of 2201 rated models, from 15 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AI for Education Pedagogy - Maths | 93.65 | Accuracy (%) | 94.4 |
| LisanBench | 0.37 | Mean Path Length / Current Maximum | 93.3 |
| AI for Education Pedagogy - Science | 92.9 | Accuracy (%) | 89.6 |
| AI for Education Pedagogy | 89.32 | Accuracy (%) | 87.4 |
| AI for Education SEND | 84.4 | Accuracy (%) | 86.7 |
| AI for Education Pedagogy - Secondary | 88.36 | Accuracy (%) | 86.2 |
| AI for Education Pedagogy - Primary | 92.49 | Accuracy (%) | 83.5 |
| cua-speedrun - OSWorld 2.0 (52 Tasks) | 68.59 | Mean verifier score (%), partial credit over the task set | 74.1 |
| DuelLab GameBench 2 - Reasoning Variants | 61.58 | GameBench2 normalized score (0–100) | 72.8 |
| Argo-Bench - Solved | 16.19 | Share of the 210 tasks solved (%), a task counting as solved | 64.8 |
| Argo-Bench | 32.87 | Mean task score (0-100) over 210 enterprise data-science tas | 64.1 |
| AI for Education Pedagogy - Technology | 82.08 | Accuracy (%) | 63.1 |
Interactive version: theaggregate.ai/model?slug=muse-wonder-medium · How It Works · Data refreshed daily, snapshot 2026-11-12.