Name: Muse Wonder Full
Author: Meta

Rank	Model	Overall score	Artificial Analysis	LiveBench	Epoch AI	ARC-AGI	Scale SEAL	Vals AI	Terminal-Bench	EQ-Bench	SimpleBench	BenchLM
1	Claude Opus 5.5 Anthropic	92.2	#1	#2	#1	#3	—	#2	#1	—	#1	#1
2	Gemini 4 Argon Google	91.9	#6	—	—	—	—	#1	—	—	—	—
3	Muse Wonder Full Meta	90.9	#3	—	—	—	—	#5	—	—	—	—
4	GPT-6 Astra OpenAI	90.7	#5	#4	#2	#1	#1	#7	#3	—	#4	#2
5	Claude Sonnet 5.5 Anthropic	90.3	#2	#21	#4	—	—	#2	#2	—	#14	#3
6	Claude Fable 5.1 Anthropic	90.3	#4	#1	#5	#6	#2	#4	#5	—	#2	#5
7	GPT-6.1 Sol OpenAI	89.8	#8	#5	#3	#2	—	#9	#3	—	#5	#4
8	Claude Opus 5 Anthropic	89.2	#9	#11	#6	#5	—	#6	#6	#1	#9	#7
9	Claude Fable 5 Anthropic	88.8	#10	#3	#9	#8	—	#8	#8	#2	—	#6
10	Muse Wonder Meta	88.5	#7	—	—	—	—	#10	—	—	—	—
11	GPT-6 Sol OpenAI	86.0	#12	#13	#7	#7	—	#13	#7	—	#17	#9
12	GPT-5.6 Sol OpenAI	85.8	#13	#8	#10	#4	—	#12	#11	#9	#25	#8
13	GPT-5.5 Pro OpenAI	85.0	—	—	#8	#11	—	—	—	—	#11	—
14	Muse Spark 1.3 Meta	84.8	#11	#5	#19	—	—	#11	#20	—	#8	—
15	GPT-5.4 Pro OpenAI	83.5	—	—	#13	#15	—	—	—	—	—	—
16	MiMo V2.6 Pro Xiaomi	83.2	#15	—	—	—	—	#14	—	—	—	—
17	GPT-5.5 OpenAI	81.7	#32	#10	#12	#10	—	—	—	#4	#21	#19
18	Gemini 3.8 Flash Google	81.6	#26	#33	#20	#8	#5	#16	#17	—	#6	#10
19	Grok 4.7 xAI	80.9	#14	#22	#41	#26	—	#16	#10	—	—	—
20	GLM-5.3 Z.ai (Zhipu)	80.0	#17	#31	#27	—	—	#19	#9	—	#24	#15
21	DeepSeek V4.1 Flash DeepSeek	79.7	#29	#8	#32	#19	—	#25	—	—	#23	#18
22	GPT-5.6 Terra OpenAI	79.5	#22	#20	#11	#14	—	#21	#15	#11	#55	#11
23	Claude Opus 4.8 Anthropic	79.3	#23	#29	#14	#20	—	#15	#14	#6	#25	#14
23	GPT-5.4 OpenAI	79.2	—	#17	#17	#18	—	—	—	#7	—	#20
25	Kimi K3 Moonshot AI	79.2	#20	#13	#15	#29	—	#29	—	#3	#35	#13
26	Grok 4.6 xAI	79.1	#18	#17	#21	#23	—	#22	#16	—	#14	#17
27	Qwen 3.8 Max Alibaba (Qwen)	78.9	#16	#16	#31	—	—	#33	#13	—	—	#12
28	Gemini 3.7 Flash Google	78.7	#30	#15	#16	#11	—	#25	#23	—	—	#16
29	MiMo V2.6 Flash Xiaomi	78.1	#36	—	—	—	—	#20	—	—	—	—
30	Step 5 StepFun	77.7	#19	—	—	—	—	#31	—	—	—	—
31	Claude Opus 4.7 Anthropic	77.5	—	#28	#25	#17	#10	—	—	#5	#28	—
32	Muse Spark 1.2 Meta	77.1	#28	#17	#33	—	—	#32	—	—	#16	#23
33	Claude Haiku 5.5 Anthropic	77.0	#21	#48	—	—	—	#18	—	—	—	—
34	Qwen 3.8 Flash Next Alibaba (Qwen)	75.8	#27	#29	—	—	—	—	—	—	—	—
35	Claude Sonnet 5 Anthropic	74.6	#34	#32	#26	—	—	#23	#21	#10	#36	#25
36	GLM-5.3 Flash Z.ai (Zhipu)	73.9	#23	#52	#47	#24	—	—	#12	—	—	—
37	Grok 4.20 xAI	73.1	—	—	#46	#25	—	—	—	—	—	—
38	GPT-6 Luna OpenAI	72.7	#35	#49	#24	#34	—	#27	#19	—	—	#24
39	Claude Opus 4.6 Anthropic	72.1	—	#39	#30	#22	#11	—	—	#13	#22	#30
39	DeepSeek V4 Pro DeepSeek	72.1	#38	#22	#29	#28	—	#36	—	#18	#53	—
41	Muse Spark 1.1 Meta	71.8	#43	#35	#38	—	—	—	—	#8	—	#22
42	GPT-5.6 Luna OpenAI	71.6	#37	#42	#23	#33	—	#24	#18	#19	#57	#21
43	DeepSeek V4 Flash DeepSeek	71.4	#40	#40	#35	#26	—	#34	—	—	#33	—
44	Muse Spark Meta	71.3	—	—	#45	—	#7	—	—	—	—	#35
44	Grok 4.5 xAI	71.2	#31	#33	#39	#37	—	#38	#21	—	#20	#26
46	Gemini 3.1 Pro Google	70.9	#49	#25	#34	#16	#3	#41	—	#21	#10	—
47	Mistral Large 4 Mistral AI	70.5	#32	#51	—	—	—	#34	—	—	—	—
48	Gemini 3.5 Flash Google	70.2	#46	#37	#36	#20	—	#37	—	#24	#12	#29
49	Gemini 3.6 Flash Google	68.9	#41	#42	#37	#29	—	—	—	—	—	#27
50	GPT-5.2 Pro OpenAI	67.1	—	—	#28	#35	—	—	—	—	#43	—

Rank	Provider	Overall score	Top model	Models
1	Anthropic	87.2	Claude Opus 5.5	31
2	OpenAI	85.2	GPT-6 Astra	54
3	Google	77.7	Gemini 4 Argon	34
4	Meta	75.4	Muse Wonder Full	26
5	xAI	69.7	Grok 4.7	13
6	DeepSeek	68.1	DeepSeek V4.1 Flash	22
7	Moonshot AI	67.8	Kimi K3	5
8	Alibaba (Qwen)	67.3	Qwen 3.8 Max	46
9	Z.ai (Zhipu)	64.9	GLM-5.3	10
10	StepFun	61.9	Step 5	1
11	Xiaomi	59.1	MiMo V2.6 Pro	3
12	Thinking Machines	41.1	Inkling	2
13	Tencent	40.0	—	2
14	MiniMax	36.1	Minimax M3	5
15	Mistral AI	34.6	Mistral Large 4	27
16	NVIDIA	29.3	Nemotron 3 Ultra 550B A55B	10
17	Amazon	18.6	—	3
18	Microsoft	13.2	Phi-4	6
19	Cohere	11.6	Command R	3
20	Inception	8.9	Mercury 2.5	1

A provider’s result on each site is that of its best model there.

Performance vs price

Top left = cheap and strong. Click a point for details.

Best value

Overall score ÷ price

Top open-weight models

Models with public weights

Fastest models

Output tokens per second

By country

Top 50 models

Agreement between leaderboards

Rank correlation on shared models (1 = identical).

How the ranking works

Collect — Official data from the 11 leaderboards below is fetched every hour. Scores marked as estimates are not used.
Match — Different spellings of the same model are merged. For variants such as reasoning effort, each site’s best result is used.
Standardize — Each site’s main metric is standardized within that site, keeping score gaps rather than just ranks.
Equate — A factor model, bridged by models listed on several sites, corrects for each site’s difficulty and model mix. Noisier sites get less weight.
No penalty for missing data — Models are scored only from the sites that rated them. Uncertainty is shown as a 90% interval. Models rated on 2+ sites are ranked.
Arena — Arena (formerly LMArena) is used for domain rankings only, not for the overall score. New models are added automatically.

Leaderboard	Rank	Published value
Artificial Analysis	#3 / 127	55.8
LiveBench	Not rated	—
Epoch AI	Not rated	—
ARC Prize (ARC-AGI)	Not rated	—
Scale AI SEAL	Not rated	—
Vals AI	#5 / 45	64.6
Terminal-Bench	Not rated	—
EQ-Bench	Not rated	—
SimpleBench	Not rated	—
BenchLM.ai	Not rated	—

Muse Wonder Full: benchmarks and ranking

Rank on each leaderboard

Domain scores