| AI models ranked by latest benchmarks

Muse Spark

Latest

Meta

•

Proprietary

# 23

Released

Apr 8, 2026

Benchmarks

# 24

Code RankedAGI

73.4%

# 29

SWEBench Pro

52.4%

# 18

SWEBench Verified

77.4%

# 22

Terminal Bench 2.0

59.0%

# 25

Agentic RankedAGI

71.6%

# 9

Code Arena

1507

# 9

Cyber Gym

43.5%

# 39

Reason RankedAGI

69.4%

# 10

HLE

42.8%

# 18

HLE w/ Tools

50.4%

# 18

GPQA Diamond

89.5%

# 6

MMMU Pro

80.4%

# 38

Math RankedAGI

71.6%

# 29

RAGI RankedAGI

65.2%

# 13

ARC AGI 2.0

42.5%

# 3

LiveCodeBench Pro

80.0%

# 15

𝜏²-Bench Telecom

91.5%

# 2

HealthBench Hard

42.8%

# 3

MedXpertQA Text

52.6%

# 2

MedXpertQA MM

78.4%

# 4

DeepSearch QA

74.8%

# 17

GDPval AA

1428

# 2

ZeroBench

33.0%

# 1

CharXiv Reasoning

86.4%

# 14

Code DesignArena

1301

Additional Details

Links

Last Updated

24 days ago