MetaAuto-detected

Muse Spark

Name: Muse Spark benchmark results
Creator: BenchLeader

Best configuration ranks #27 of 610 on the BenchLeader Index at 65.4 ±3.3. Last measured 2 Sept 2026. Released 8 Apr 2026.

Blended price: –
Output speed: –
First answer: –
Context: 262k

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

Overall index65
Reasoning70
Coding66
Agents & tools66
Maths56
Knowledge65
Instruction following80
Human preference69
Multimodal66
Long context66
Composite68

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. “Default” means the publisher did not say which setting was used.

Effort	Index	Rank	Speed	First answer	Chat reply cost	Categories
thinking	–	–	–	–	–
defaultbest	65.4	#27	–	–	–	Agents & tools 66 · Coding 66 · Composite 68 · Human preference 69 · Instruction following 80 · Knowledge 65 · Long context 66 · Maths 56 · Multimodal 66 · Reasoning 70

Benchmark results

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	thinking	default	Source
GPQA Diamond	–	89.8%#42	Epoch AI Benchmarking Hub
Humanity's Last Exam	–	40.6%#4	Scale AI / CAIS
LMArena Hard Prompts	–	1505#18	LMArena
GPQA Diamond (AA)not in index	–	88.4%#89	Artificial Analysis
Humanity's Last Exam (AA)not in index	–	40.7%#52	Artificial Analysis
GPQA Diamond (Vals)not in index	–	89.7%#28	Vals AI

Coding

Benchmark	thinking	default	Source
SciCode	–	51.5%#52	SciCode
LMArena Coding	–	1526#18	LMArena
SWE-bench (Vals)not in index	–	74.4%#46	Vals AI
SWE-Bench Pro	–	55.0%#3	Scale AI SEAL

Agents & tools

Benchmark	thinking	default	Source
Terminal-Bench Hard	–	45.5%#32	Artificial Analysis
τ²-Bench Telecom (AA)not in index	–	91.5%#58	Artificial Analysis
MCP Atlas	–	82.2%#7	Scale AI SEAL

Maths

Benchmark	thinking	default	Source
OTIS Mock AIME	–	88.9%#60	Epoch AI Benchmarking Hub
ProofBench	–	17.0%#40	Vals AI
AIME (Vals)	–	96.9%#2	Vals AI

Knowledge

Benchmark	thinking	default	Source
AA-Omniscience	–	7.2#67	Artificial Analysis
MMLU-Pro	–	87.3%#30	Vals AI
LegalBench	–	84.2%#34	Vals AI
CorpFin	–	65.1%#33	Vals AI
TaxEval	–	77.7%#3	Vals AI
PRBench Finance	–	52.4%#4	Scale AI SEAL
PRBench Legal	–	52.3%#3	Scale AI SEAL
MultiNRC	–	59.0%#5	Scale AI SEAL

Instruction following

Benchmark	thinking	default	Source
IFBench	–	75.9%#24	Artificial Analysis
MultiChallenge	–	75.5%#1	Scale AI SEAL
TutorBench	–	68.5%#1	Scale AI SEAL

Human preference

Benchmark	thinking	default	Source
LMArena Text	–	1488#13	LMArena

Multimodal

Benchmark	thinking	default	Source
LMArena Vision	–	1306#9	LMArena
MMMU-Pro	–	80.5%#31	Artificial Analysis
MMMU-Pro (official)not in index	80.4%#3	–	MMMU

Long context

Benchmark	thinking	default	Source
AA-LCR	–	78.0%#82	Artificial Analysis

Composite

Benchmark	thinking	default	Source
Epoch Capabilities Indexnot in index	–	152.1#35	Epoch AI Benchmarking Hub
AA Intelligence Index	–	31.3#73	Artificial Analysis

Where it wins

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