Agentic Tool Use on Tau2-Telecom
72.8Avg@8LongCat-Flash-Lite
Evaluation Results
| Method | Links | |
|---|---|---|
| LongCat-Flash-LiteArchitecture=MoE + NE, # Total Params=68.5B, # Activated Params=2.9B~4.5B2026.01 | 72.8 | |
| LongCat-Next2026.03 | 62.06 | |
| Gemini 2.5 Flash-Lite2026.01 | 21.93 | |
| Kimi-Linear-48B-A3BArchitecture=MoE, # Total Params=48B, # Activated Params=3B2026.01 | 15.68 | |
| Kimi-Linear-48B-A3B2026.03 | 15.68 | |
| Qwen3-Next-80B-A3B-InstructArchitecture=MoE, # Total Params=80B, # Activated Params=3B2026.01 | 13.2 | |
| Qwen3-Next-80B-A3B-Instruct2026.03 | 13.2 | |
| Qwen3-Omni-A3B-Instruct2026.03 | 4.39 |