Leaderboard
Muse Wonder Full
| Released | 2026-10-12 |
|---|
| Price per 1M | Not listed |
|---|
| Overall score | 62.8 · rank 13 |
|---|
| Context window | 1,048,576 tokens |
|---|
| Provider | Meta |
|---|
| Weights | Proprietary |
|---|
Scores
Buzz
| Posts on X | About 28 a day |
|---|
| Mood | 11% positive, 0% negative |
|---|
| Hacker News | 1 mention this week |
|---|
| Checked | 2026-11-11 |
|---|
- Congrats to @GoogleDeepMind on Gemini 4 Carbon, now #1 on SWE-Marathon v1.1, our long-horizon coding benchmark.
53.8% pass rate, ahead of Muse Wonder Full (53.1%), Claude Opus 5.5 (46.3%) and GPT-6 Astra (42.5%).
88% cheaper per solved task than Opus 5.5. https://t.co/LAoaySs4…@abundant_labs · 282 likes
- - Most frontier models today have too many parameters. A lot of these params encode knowledge you wouldn't need 99.9% of the time but you still pay for it on every inference call. (MoE doesn't fully fix this).
- Distillation really solves this problem. Muse Ember is a perfect e…@harris_p10 · 70 likes
- Claude Fable 5.2 lands 2nd on GDPx legal, behind Muse Wonder Full, up 4.6 points on Fable 5.1. https://t.co/KXpKZ5er5D@chakra_ai · 21 likes
Configurations
Evidence
Coding 60.9 ±2.2
Agents 62.8 ±1.0
Reasoning 64.6 ±3.4
Math 64.0 ±2.7
Knowledge 64.3 ±6.5
Frontend 61.3 ±0.7
Names matched to this model
muse-wonder-full
Compare Muse Wonder Full with other models