BenchLeader

Muse Spark vs Muse Spark 1.3

Verdict
  • Muse Spark leads on quality: 65.4 vs 63.9.
  • Muse Spark is stronger in agents & tools, human preference, instruction following, maths, reasoning.
  • Muse Spark 1.3 (xhigh) is stronger in coding, composite, knowledge, long context, multimodal.
MetricMuse SparkMuse Spark 1.3 (xhigh)
BenchLeader Index65.463.9
Agents & tools score66.360.4
Coding score66.070.8
Composite score68.478.6
Human preference score69.0
Instruction following score79.6
Knowledge score64.575.0
Long context score65.568.1
Maths score56.3
Multimodal score65.966.6
Reasoning score70.0
Blended price $/M$2.00
Output speed193 tok/s
Time to first answer40.0 s
Context window262k1M
GPQA Diamond89.8%
OTIS Mock AIME88.9%
Humanity's Last Exam40.6%
SciCode51.5%
ProofBench17.0%
Epoch Capabilities Index152.1
LMArena Text1488
LMArena Hard Prompts1505
LMArena Coding1526
LMArena WebDev1625
LMArena Vision1306
LiveBench81.6%
LiveBench Reasoning89.7%
LiveBench Coding81.1%
LiveBench Agentic Coding64.1%
LiveBench Mathematics96.0%
LiveBench Data Analysis79.6%
LiveBench Language82.8%
AA Intelligence Index31.345.2
IFBench75.9%
AA-LCR78.0%83.0%
MMMU-Pro80.5%82.0%
AA-Omniscience7.223.1
Terminal-Bench Hard45.5%
GPQA Diamond (AA)88.4%94.1%
Humanity's Last Exam (AA)40.7%47.5%
SciCode (AA)59.7%
τ²-Bench Telecom (AA)91.5%
AIME (Vals)96.9%
MMLU-Pro87.3%
LegalBench84.2%
CorpFin65.1%
TaxEval77.7%
Terminal-Bench 2.1 (Vals)72.3%
SWE-bench (Vals)74.4%
GPQA Diamond (Vals)89.7%
Vals Index60.3
SWE-Bench Pro55.0%
MCP Atlas82.2%
MultiChallenge75.5%
PRBench Finance52.4%
PRBench Legal52.3%
MultiNRC59.0%
TutorBench68.5%

Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.