List your company

Reasoning on HLE OOD

38.6AccuracyMTRouter
7.19215.34623.531.654Apr 26, 2026

Evaluation results

MethodLinks
MTRouterRouting Level=Multi-Turn Routers (turn-level)2026.0438.631.2
LLM RouterRouting Level=Multi-Turn Routers (turn-level)2026.043635.6
Router-R1Routing Level=Multi-Turn Routers (turn-level)2026.0435.160.7
GPT-5Routing Level=Single-Model Baselines2026.0434.865.3
OpenRouterRouting Level=Multi-Turn Routers (turn-level)2026.0434154.3
EmbedLLMRouting Level=Single-Turn Routers (episode-level)2026.0433.656.8
AvengersProRouting Level=Single-Turn Routers (episode-level)2026.0430.633.3
DeepSeek-V3.2Routing Level=Single-Model Baselines2026.0428.722.2
Random RouterRouting Level=Multi-Turn Routers (turn-level)2026.0423.814.7
Kimi-K2Routing Level=Single-Model Baselines2026.0420.19.8
RouterDCRouting Level=Single-Turn Routers (episode-level)2026.0417.913.5
GPT-OSS-120BRouting Level=Single-Model Baselines2026.0411.42.1
MiniMax-M2Routing Level=Single-Model Baselines2026.048.99.8
Gemini-2.5-Flash-LiteRouting Level=Single-Model Baselines2026.048.42.1