Buyer Negotiation on AGORABENCH (test)
1.414Vanilla Single MERITgpt-4o
Evaluation Results
| Method | Links | ||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| gpt-4oFramework=ReAct2026.02 | 1.414 | 86.7 | 1.342 | 86.4 | 1.275 | 84.6 | 1.139 | 78.9 | 1.211 | 84.6 | 1.188 | 84.4 | 1.032 | 78.6 | 1.281 | 83.9 | 1.342 | 87.4 | 1.247 | 83.9 | |
| gpt-3.5-turboFramework=ReAct2026.02 | 1.322 | 94.4 | 1.366 | 90 | 1.344 | 90.5 | 1.159 | 81.3 | 1.159 | 95.8 | 1.254 | 90.7 | 1.177 | 84.6 | 1.14 | 78.2 | 1.395 | 91 | 1.257 | 88.5 | |
| gemini-1.5-proFramework=ReAct2026.02 | 1.202 | 72 | 1.394 | 75.8 | 1.502 | 93.6 | 1.35 | 74.6 | 1.324 | 66.7 | 1.522 | 89.4 | 1.308 | 94.2 | 1.3 | 68 | 1.525 | 88.8 | 1.381 | 80.3 | |
| gpt-4o-miniFramework=ReAct2026.02 | 1.164 | 77.1 | 1.228 | 74.4 | 1.402 | 89.7 | 1.23 | 79.6 | 1.381 | 89.9 | 1.105 | 78.9 | 1.376 | 88.5 | 0.973 | 56 | 1.305 | 76.7 | 1.24 | 79 | |
| gemini-1.5-flashFramework=ReAct2026.02 | 0.875 | 61.8 | 1.005 | 66.3 | 1.304 | 87.6 | 0.951 | 61.8 | 1.11 | 65 | 0.988 | 71.8 | 1.056 | 75.4 | 0.826 | 54.4 | 1.284 | 82.6 | 1.044 | 69.6 |