Compare Models
Muse Spark 1.2

Release Date: Aug 5, 2026

Developer Meta  🇺🇸
Context Window 1M
Max Output Tokens 131k
Token Costs (in/out) $1.25/4.25
Weights Private
Input Modalities

Accuracy

57.05 % ± 1.12

Cost / Test (Vals Index)

$ 1.662

Latency

16 min 49 s

Vals Index
BenchmarksAccuracyRankings
Vals Index

0.0%

±1.12
18/58
Code Migration

0.0%

±4.02
27/61
EMB

0.0%

±3.13
28/58
Finance Agent (v2)

0.0%

±0.28
2/61
Legal Research Bench

0.0%

±3.45
12/61
MedCode

0.0%

±2.19
17/93
MedScribe

0.0%

±1.96
3/95
MortgageTax

0.0%

±0.93
43/98
ProofBench v1.1

0.0%

±4.95
19/32
SAGE

0.0%

±3.44
27/81
Public Benefits Bench v1.1

0.0%

±1.21
6/36
TaxEval v2

0.0%

±0.76
1/145
Vibe Code Bench v1.1

0.0%

±3.31
13/96
BioMysteryBench

0.0%

±1.33
8/12
Harvey's Legal Agent Benchmark

0.0%

±3.78
1/62
IOI

0.0%

±0.87
25/28
LegalBench

0.0%

±0.45
20/145
MMLU Pro

0.0%

±0.32
20/138
MMMU Pro

0.0%

±0.83
22/93
SkillsBench

0.0%

±4.42
17/35
SWE-bench

0.0%

±1.52
14/88
Terminal-Bench 4.0

0.0%

±2.67
21/28
Proprietary BenchmarksAcademic BenchmarksIndustry Partners
Vals
Default Provider : Meta
Temperature: 1
Top P: Default
Top K: Default
Max Output Tokens: 131,072
Reasoning Effort: xhigh