Function Calling for Web Search on BFCL Web Search
86.5Task Completion Score (TCS)Gemini-3-flash
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini-3-flashAverage Steps (Stp)=10.2, Input Tokens (Tok In)=75.4, Output Tokens (Tok Out)=15.5, Execution Time=90.22026.05 | 86.5 | |
| GPT-5-miniAverage Steps (Stp)=11.3, Input Tokens (Tok In)=65.6, Output Tokens (Tok Out)=6.5, Execution Time=116.82026.05 | 83.5 | |
| DeepSeek-V3.2Average Steps (Stp)=7.1, Input Tokens (Tok In)=31.5, Output Tokens (Tok Out)=0.4, Execution Time=8.52026.05 | 80 | |
| Qwen3-235B-A22BAverage Steps (Stp)=4.7, Input Tokens (Tok In)=17.5, Output Tokens (Tok Out)=0.3, Execution Time=3.82026.05 | 68 |