Muse Wonder Full
Meta AI
Muse Wonder Full is a large language model developed by Meta AI. It is designed for complex, autonomous tasks.
Token pricing and context
- Input
- Not reported
- Cached input
- Not reported
- Output
- Not reported
- Context window
- 1000000
Benchmark results
MM1.0
Overall score: 84.5 · Scale 0–100 · Higher is better
- Reasoning: 95
- Coding: 84
- Chat: 90
- Long Context: 100
- Vibe: 90
- Accuracy: 100
- General Knowledge: 37.5
Evaluated: 2026-10-18
Record exact model version, reasoning setting, temperature, tools, and evaluation setup.
Hallucimeter 1.0
Overall score: 17 · Scale 0–100 · Lower is better
- Hallucinations: 17
Evaluated: 2026-10-18
Record exact model version, reasoning setting, temperature, tools, and evaluation setup.
Updated 2026-10-18. Read benchmark methodology.