Task Efficient Routing on Data scenario
0.0043CostForce Weak
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Force Weak2026.01 | 0.0043 | — | 90.8 | |
| Force WeakModel=qwen2026.01 | 0.0136 | 70.2 | — | |
| CASTER2026.01 | 0.0255 | — | 45.3 | |
| CASTERModel=qwen2026.01 | 0.0344 | 24.6 | — | |
| Force WeakModel=openai2026.01 | 0.0452 | 96.1 | — | |
| Force StrongModel=qwen2026.01 | 0.0456 | — | — | |
| Force Strong2026.01 | 0.0466 | — | — | |
| CASTERModel=deepseek2026.01 | 0.1332 | 39.3 | — | |
| Force StrongModel=deepseek2026.01 | 0.2193 | — | — | |
| Force WeakModel=deepseek2026.01 | 0.2268 | -3.4 | — | |
| Force WeakModel=gemini2026.01 | 0.3009 | 74.2 | — | |
| CASTERModel=gemini2026.01 | 0.3258 | 47.6 | — | |
| CASTERStrategy=Dynamic Routing2026.01 | 0.41 | — | 38.4 | |
| CASTERModel=openai2026.01 | 0.4533 | 60.6 | — | |
| Force WeakModel=claude2026.01 | 0.4641 | 85.6 | — | |
| FrugalGPT (Cascade)Strategy=Cascade2026.01 | 0.66 | — | — | |
| CASTERModel=claude2026.01 | 0.9186 | 71.5 | — | |
| Force StrongModel=openai2026.01 | 1.1514 | — | — | |
| Force StrongModel=gemini2026.01 | 1.8052 | — | — | |
| Force StrongModel=claude2026.01 | 3.2237 | — | — |