Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Model Routing research benchmarks – Page 2 · SOTA2 Research

Model Routing

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

60.9CPT (80%)

4

Mar 12, 2026

68.24CPT (80%)

4

Mar 12, 2026

MMLU-Redux (ID)

52.93CPT (80%)

4

Mar 12, 2026

52.61CPT (80%)

4

Mar 12, 2026

GPQA Diamond (ID)

62.69CPT (80%)

4

Mar 12, 2026

FRAMES (ID queries)

69.41CPT (85%) Score

4

Mar 12, 2026

MATH-500 (ID queries)

64.97CPT (85%)

4

Mar 12, 2026

AIME ID queries

80.84CPT (85%)

4

Mar 12, 2026

LSAT ID queries

70.35CPT (85%)

4

Mar 12, 2026

MMLU-Pro (ID queries)

75.91CPT Score (85%)

4

Mar 12, 2026

MMLU-Redux (ID queries)

63.77CPT (85%)

4

Mar 12, 2026

MMLU (ID queries)

64.46CPT (85%) Score

4

Mar 12, 2026

GPQA Diamond (ID queries)

71.53CPT (85%)

4

Mar 12, 2026

86.34CPT (95%)

4

Mar 12, 2026

90CPT (95%)

4

Mar 12, 2026

91.32CPT (95%)

4

Mar 12, 2026

86.14CPT Score (95%)

4

Mar 12, 2026

86.42CPT Accuracy (95%)

4

Mar 12, 2026

90.78CPT (95%)

4

Mar 12, 2026

EmbedLLM (test)

67.2Accuracy

4

Feb 26, 2026

89.3CPT (95%)

3

Mar 12, 2026

92.44CPT (95%)

2

Mar 12, 2026

Page 2 of 2

Popular datasets

MMLU MATH AIME GPQA LSAT

Follow for updates

Join SOTA2 Contact