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

LLM Routing research benchmarks – Page 3 · SOTA2 Research

LLM Routing

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

RouteLLM in-distribution

92.42HV Score

6

Jun 5, 2026

SWE-Bench Verified n=500

OpenRouter Auto

92.1QR (%)

6

May 19, 2026

Multilingual production VS Code Chat telemetry (90 days) (5,001-query eval set)

100Routing Success Rate

6

May 19, 2026

Oracle (upper bound)

0.077AUDC

6

Apr 20, 2026

Oracle (upper bound)

77.55AUDC

6

Apr 20, 2026

RouterBench held-out (test)

91.3Accuracy

6

Feb 27, 2026

LLM Routing Systems

40Latency Overhead (ms)

5

May 19, 2026

LogicGrid, MGSM, and CommonGen Unseen Average Phase-2 (test)

78Average Accuracy

4

Apr 28, 2026

CommonGen Phase-2 (test)

82Accuracy

4

Apr 28, 2026

MGSM Phase-2 (test)

92Accuracy

4

Apr 28, 2026

LogicGrid Phase-2 (test)

60Accuracy

4

Apr 28, 2026

MMLU Mathematics and History Error + Cost setting

31.2Deferral Ratio (7B, Str)

3

May 1, 2026

MMLU Mathematics and History Error Only setting

99.67B (Str) Deferral Ratio

3

May 1, 2026

96Accuracy

3

Feb 27, 2026

MMLU-Pro non-STEM

71AUDC

3

Feb 26, 2026

—

—Accuracy

0

Feb 26, 2026

Page 3 of 3

Popular datasets

MMLU MATH RouterBench In-domain datasets MedMCQA

Follow for updates

Join SOTA2 Contact