MetaReasoning modelCompare with another model

Muse Spark

Reasoning effort

Muse Spark is a Meta proprietary reasoning model, released 8 Apr 2026. Its best configuration ranks #50 of 775 on the BenchLeader Index at 64.4 ±3.1, in the upper half. The ± is the point: 124 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 instruction following (78) and lowest in maths (59). It has been measured at 2 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 19 Oct 2026.

Blended price: –
Output speed: –
First answer: –
Context: 262k
Full answer: –
Cost per run: –
Released: 8 Apr 2026

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

Overall index64
Reasoning65
Coding65
Agents & tools66
Maths59
Knowledge63
Instruction following78
Human preference69
Multimodal65
Long context64
Composite65

Versions

Meta has shipped 4 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.

Model	Released	Index	Rank
Muse Spark 1.3max	2 Sept 2026	67.2	#26
Muse Spark 1.2xhigh	5 Aug 2026	63.2	#69
Muse Spark 1.1	9 Jul 2026	62.9	#75
Muse Sparkthis page	8 Apr 2026	64.4	#50

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
thinking	–	–	–	–	–
not statedbest	64.4	#50	–	–	–	Agents & tools 66 · Coding 65 · Composite 65 · Human preference 69 · Instruction following 78 · Knowledge 63 · Long context 64 · Maths 59 · Multimodal 65 · Reasoning 65

How its index has moved

8 Sept 2026 to 20 Oct 2026

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	not stated	Source
GPQA Diamond	89.8%#50	Epoch AI Benchmarking Hub
Humanity's Last Exam	40.6%#7	Scale AI / CAIS
Arena Hard Prompts	1506#22	Arena (LMArena)
GPQA Diamond (AA)not in index	88.4%#89	Artificial Analysis
Humanity's Last Exam (AA)not in index	40.7%#82	Artificial Analysis
GPQA Diamond (Vals)not in index	89.7%#28	Vals AI
CritPt	11.3%#111	Artificial Analysis

Coding

Benchmark	not stated	Source
SciCode	51.5%#97	SciCode
Arena Coding	1529#24	Arena (LMArena)
SWE-bench (Vals)not in index	74.4%#46	Vals AI
SWE-Bench Pro	55.0%#3	Scale AI SEAL
Vibe Code Bench v1.1not in index	19.7%#82	Vals AI
SWE-Bench Pro (private)not in index	55.0%#3	Scale AI SEAL
SWE Atlas: Codebase QnAnot in index	24.2%#17	Scale AI SEAL
SWE Atlas: Test Writingnot in index	31.1%#16	Scale AI SEAL
Terminal-Bench 2.1 (AA)not in index	62.2%#90	Artificial Analysis

Agents & tools

Benchmark	not stated	Source
Terminal-Bench Hard	45.5%#34	Artificial Analysis
τ²-Bench Telecom (AA)not in index	91.5%#57	Artificial Analysis
MCP Atlas	82.2%#9	Scale AI SEAL
GDPval-AA v2.1not in index	24.9%#161	Artificial Analysis
Terminal-Bench 2.0 (Vals)not in index	59.5%#7	Vals AI

Maths

Benchmark	not stated	Source
OTIS Mock AIME	88.9%#74	Epoch AI Benchmarking Hub
ProofBench	17.0%#55	Vals AI
AIME (Vals)	96.9%#2	Vals AI
Arena Maths	1467#56	Arena (LMArena)

Knowledge

Benchmark	not stated	Source
AA-Omniscience	7.2#99	Artificial Analysis
MMLU-Pro	87.3%#30	Vals AI
LegalBench	84.2%#38	Vals AI
CorpFin	65.1%#33	Vals AI
TaxEval	77.7%#3	Vals AI
PRBench Finance	52.4%#6	Scale AI SEAL
PRBench Legal	52.3%#6	Scale AI SEAL
MultiNRCpossible contamination	59.0%#5	Scale AI SEAL
CaseLaw v2not in index	63.1%#17	Vals AI
MedCodenot in index	51.3%#16	Vals AI
MedScribenot in index	85.9%#25	Vals AI
AA-Omniscience: accuracynot in index	49.6%#64	Artificial Analysis
AA-Omniscience: non-hallucinationnot in index	15.8%#330	Artificial Analysis

Instruction following

Benchmark	not stated	Source
IFBench	75.9%#24	Artificial Analysis
MultiChallenge	75.5%#2	Scale AI SEAL
TutorBench	68.5%#1	Scale AI SEAL
Arena Instruction Followingnot in index	1464#45	Arena (LMArena)

Human preference

Benchmark	not stated	Source
Arena Text	1489#16	Arena (LMArena)
Arena Creative Writingnot in index	1465#23	Arena (LMArena)
Arena Multi-turnnot in index	1494#18	Arena (LMArena)
Arena Longer Queriesnot in index	1476#52	Arena (LMArena)

Multimodal

Benchmark	thinking	not stated	Source
Arena Visioneffort not stated	–	1306#15	Arena (LMArena)
MMMU-Proeffort not stated	–	80.5%#45	Artificial Analysis
MMMU-Pro (official)not in index	80.4%#3	–	MMMU
MMMU-Pro (Vals)not in indexeffort not stated	–	87.4%#16	Vals AI

Long context

Benchmark	not stated	Source	Trend
AA-LCR	78.0%#120	Artificial Analysis
Arena Documentnot in index	1467#20	Arena (LMArena)

Composite

Benchmark	not stated	Source	Trend
Epoch Capabilities Indexnot in index	152.0#50	Epoch AI Benchmarking Hub
AA Intelligence Index v4.3.2	31.3#114	Artificial Analysis

Safety & honesty

Benchmark	not stated	Source	Trend
FORTRESSnot in index	20.2%#39	Scale AI SEAL

Where it wins

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