| AI models ranked by latest benchmarks

Muse Spark

Latest

Meta

•

Proprietary

# 10

Released

Apr 8, 2026

Benchmarks

# 11

Code RankedAGI

80.8%

# 20

SWEBench Pro

52.4%

# 15

SWEBench Verified

77.4%

# 20

Terminal Bench 2.0

59.0%

# 11

Agentic RankedAGI

73.1%

# 6

Code Arena

1510

# 10

Cyber Gym

43.5%

# 26

Reason RankedAGI

69.9%

# 7

HLE

42.8%

# 14

HLE w/ Tools

50.4%

# 17

GPQA Diamond

89.5%

# 5

MMMU Pro

80.4%

# 31

Math RankedAGI

72.6%

# 14

RAGI RankedAGI

67.3%

# 12

ARC AGI 2.0

42.5%

# 3

LiveCodeBench Pro

80.0%

# 14

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%

# 14

GDPval AA

1428

# 2

ZeroBench

33.0%

# 1

CharXiv Reasoning

86.4%

# 6

Code DesignArena

1318

Additional Details

Links

Last Updated

2 days ago