MetaReasoning modelCompare with another model

Muse Wonder

Reasoning effort

Muse Wonder is a Meta proprietary reasoning model, released 12 Oct 2026. Its best configuration (max reasoning effort) ranks #23 of 775 on the BenchLeader Index at 67.5 ±4.7, in the top 25. The ± is the point: 84 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in composite (88) and lowest in agents & tools (56). At $8.00 per million tokens blended it is among the most expensive ranked models. Output speed of 161 tokens per second puts it in the fastest quarter, with a first answer in 44.5 s. It has been measured at 7 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 19 Oct 2026.

Blended price: $8.00/M; $4.00 in · $20.00 out
Output speed: 161 tok/s; measured by Artificial Analysis
First answer: 45 s; first token 3.02 s
Context: 1.0M
Full answer: 48 s; median, reasoning included
Cost per run: $4.52; one full Intelligence Index run
Released: 12 Oct 2026

How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.

Overall index68
Reasoning82
Coding66
Agents & tools56
Maths68
Knowledge73
Human preference71
Long context65
Composite88

Reasoning-effort configurations

The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.

Effort	Index	Rank	Speed	First answer	Chat reply cost	Categories
minimal	–	–	77 tok/s	3.02 s	$0.0076
low	–	–	77 tok/s	3.02 s	$0.0076
medium	–	–	77 tok/s	3.02 s	$0.0076
high	–	–	77 tok/s	3.02 s	$0.0076
xhigh	–	–	77 tok/s	3.02 s	$0.0076
maxbest	67.5	#23	161 tok/s	45 s	$0.0076	Agents & tools 56 · Coding 66 · Composite 88 · Human preference 71 · Knowledge 73 · Long context 65 · Maths 68 · Reasoning 82
not stated	63.6	#60	77 tok/s	3.02 s	$0.0076	Agents & tools 79 · Composite 61 · Knowledge 76 · Maths 71

How its index has moved

13 Oct 2026 to 21 Oct 2026

max
not stated

The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.

Benchmark results

See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Reasoning

Benchmark	max	not stated	Source
Arena Hard Prompts	1527#6	–	Arena (LMArena)
LiveBench 2.0 Reasoningnot in indexeffort not stated	–	90.6%#11	LiveBench
LiveBench 1.0 Reasoningnot in index	90.6%#11	–	LiveBench
Humanity's Last Exam (AA)not in index	50.1%#24	–	Artificial Analysis
CritPt	28.3%#26	–	Artificial Analysis
MysteryMechanismnot in index	27.5%#15	–	Vals AI
Mystery Game Puzzlesnot in index	51.0%#13	–	Epoch AI Benchmarking Hub
LMCAnot in indexeffort not stated	–	56.8%#41	Epoch AI Benchmarking Hub
DTBenchnot in indexeffort not stated	–	95.2%#43	Epoch AI Benchmarking Hub

Coding

Benchmark	minimal	low	medium	high	xhigh	max	not stated	Source
SciCode	–	–	–	–	–	60.3%#10	–	SciCode
Arena Coding	–	–	–	–	–	1548#5	–	Arena (LMArena)
Arena WebDev	–	–	–	–	–	1652#16	–	Arena (LMArena)
LiveBench 2.0 Codingnot in indexeffort not stated	–	–	–	–	–	–	86.2%#19	LiveBench
LiveBench 1.0 Codingnot in index	–	–	–	–	–	80.7%#18	–	LiveBench
SciCode (AA)not in index	–	–	–	–	–	60.3%#11	–	Artificial Analysis
IOI	–	–	–	–	–	66.9%#15	–	Vals AI
Code Migrationnot in index	–	–	–	–	–	62.5%#6	–	Vals AI
Vibe Code Bench 1-100not in index	–	–	–	–	–	26.6%#6	–	Vals AI
Vibe Code Bench v1.1not in index	–	–	–	–	–	83.1%#19	–	Vals AI
CursorBenchnot in index	34.2%#52	41.7%#32	45.2%#23	46.1%#21	48.2%#16	50.4%#12	–	Cursor
Terminal-Bench 4.0 (AA)not in index	–	–	–	–	–	52.0%#15	–	Artificial Analysis

Agents & tools

Benchmark	max	not stated	Source
Arena Agent	3.8#12	–	Arena (LMArena)
LiveBench Long Agenticnot in indexeffort not stated	–	53.2%#10	LiveBench
LiveBench Browsernot in indexeffort not stated	–	72.0%#20	LiveBench
LiveBench Agentic Codingnot in index	62.7%#9	–	LiveBench
HiL-Bencheffort not stated	–	79.7%#1	Scale AI SEAL
GDPval-AA v2.1not in index	63.5%#5	–	Artificial Analysis
CUA-benchnot in index	7.2%#6	–	Vals AI
CyberBenchnot in index	80.4%#1	–	Vals AI
Finance Agent v2not in index	56.6%#18	–	Vals AI
Harvey's Legal Agent Benchmarknot in index	23.8%#2	–	Vals AI
Legal Research Benchnot in index	50.5%#6	–	Vals AI
Tax Agent Benchnot in index	36.3%#11	–	Vals AI
Terminal-Bench 4.0 (Vals)not in index	49.5%#9	–	Vals AI
Terminal-Bench Sciencenot in index	22.9%#10	–	Vals AI
Vending-Bench 2not in indexeffort not stated	–	9432.2#9	Andon Labs
AutomationBenchnot in index	68.1%#8	–	Artificial Analysis
GDP.pdf (AA)not in index	30.6%#8	–	Artificial Analysis
Harvey LAB-AA v1.1not in index	10.0%#2	–	Harvey
AA-Briefcase v1.1not in index	1659#9	–	Artificial Analysis

Maths

Benchmark	max	not stated	Source
FrontierMath Tiers 1–3	82.1%#15	–	Epoch AI Benchmarking Hub
FrontierMath Tier 4	70.7%#20	–	Epoch AI Benchmarking Hub
OTIS Mock AIME	99.4%#15	–	Epoch AI Benchmarking Hub
ProofBencheffort not stated	–	84.0%#11	Vals AI
LiveBench Mathematicsnot in index	96.5%#8	–	LiveBench

Knowledge

Benchmark	max	not stated	Source
LiveBench 2.0 Data Analysisnot in indexeffort not stated	–	81.6%#4	LiveBench
LiveBench Researchnot in indexeffort not stated	–	88.0%#2	LiveBench
LiveBench 1.0 Data Analysisnot in index	81.6%#4	–	LiveBench
AA-Omniscience	25.5#50	–	Artificial Analysis
PRBench Financeeffort not stated	–	58.7%#2	Scale AI SEAL
PRBench Legaleffort not stated	–	58.4%#3	Scale AI SEAL
BioMysteryBenchnot in index	67.0%#17	–	Vals AI
Excel Modeling Benchmarknot in index	67.1%#18	–	Vals AI
Public Benefits Benchnot in index	70.2%#6	–	Vals AI
EBR-benchnot in index	31.4%#12	–	Epoch AI Benchmarking Hub
GDP.pdfnot in index	30.8%#4	–	Epoch AI Benchmarking Hub
AA-Omniscience: accuracynot in index	41.1%#113	–	Artificial Analysis
AA-Omniscience: non-hallucinationnot in index	73.5%#32	–	Artificial Analysis

Instruction following

Benchmark	max	Source
LiveBench Languagenot in index	82.9%#21	LiveBench
LiveBench Instruction Followingnot in index	73.0%#17	LiveBench
Arena Instruction Followingnot in index	1492#11	Arena (LMArena)

Human preference

Benchmark	max	Source
Arena Text	1509#2	Arena (LMArena)
Arena Creative Writingnot in index	1485#8	Arena (LMArena)
Arena Multi-turnnot in index	1530#2	Arena (LMArena)
Arena Longer Queriesnot in index	1500#14	Arena (LMArena)

Multimodal

Benchmark	not stated	Source	Trend
LiveBench Designnot in index	75.1%#28	LiveBench

Long context

Benchmark	max	Source	Trend
AA-LCR	80.7%#64	Artificial Analysis

Composite

Benchmark	max	not stated	Source
Epoch Capabilities Indexnot in indexeffort not stated	–	161.8#12	Epoch AI Benchmarking Hub
LiveBench 2.0effort not stated	–	78.1%#11	LiveBench
LiveBench 1.0not in index	81.2%#8	–	LiveBench
AA Intelligence Index v4.3.2	52.4#10	–	Artificial Analysis
Vals Indexnot in index	64.9#6	–	Vals AI

Where it wins

Benchmarks where this configuration ranks in the top five of every configuration measured.

Where to run it

Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.

Provider	Speed	First token	Input $/M	Output $/M	Context	Quantisation
Meta	87 tok/s	4.53 s	$4.00	$20.00	1.0M	–

What a task costs

Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.200 per 1M. Reasoning tokens are not modelled.

Workload	Tokens in / out	Cost	With caching	Time
Chat reply	400 / 300	$0.0076	$0.0065	46.4 s
Summarize a 30-page report	12,000 / 600	$0.060	$0.026	48.2 s
Code edit	6,000 / 1,500	$0.054	$0.037	53.8 s
Agentic coding session	60,000 / 4,000	$0.320	$0.149	1.2 min
Structured extraction	2,000 / 200	$0.012	$0.0063	45.8 s