Agentic Tool Use on Tau2 Retail
73.68Avg@8LongCat-Next
Evaluation results
| Method | Links | |
|---|---|---|
| LongCat-Next2026.03 | 73.68 | |
| LongCat-Flash-LiteArchitecture=MoE + NE, # Total Params=68.5B, # Activated Params=2.9B~4.5B2026.01 | 73.1 | |
| Qwen3-Next-80B-A3B-InstructArchitecture=MoE, # Total Params=80B, # Activated Params=3B2026.01 | 57.3 | |
| Qwen3-Next-80B-A3B-Instruct2026.03 | 57.3 | |
| Qwen3-Omni-A3B-Instruct2026.03 | 40.8 | |
| Gemini 2.5 Flash-Lite2026.01 | 37.5 | |
| Kimi-Linear-48B-A3BArchitecture=MoE, # Total Params=48B, # Activated Params=3B2026.01 | 18.86 | |
| Kimi-Linear-48B-A3B2026.03 | 18.86 |