Invoice Reconciliation on Synthetic invoice reconciliation dataset (test)
100SR (%)gpt-4.1
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| gpt-4.1Platform=OpenAI2026.01 | 100 | — | — | |
| qwen2.5:32bPlatform=RTX A60002026.01 | 100 | — | — | |
| gpt-4.1-miniPlatform=OpenAI2026.01 | 99.5 | — | — | |
| gpt-4o-miniPlatform=OpenAI2026.01 | 99.3 | — | — | |
| claude-3-7-sonnetPlatform=Anthropic2026.01 | 99.3 | — | — | |
| gpt-4oPlatform=OpenAI2026.01 | 98.8 | — | — | |
| claude-3-5-sonnetPlatform=Anthropic2026.01 | 97.8 | — | — | |
| qwen2.5:14bPlatform=RTX A60002026.01 | 96.6 | — | — | |
| qwen2.5:72bPlatform=RTX A60002026.01 | 95.1 | — | — | |
| gpt-4.1-nanoPlatform=OpenAI2026.01 | 85.7 | — | — | |
| qwen2.5:7bPlatform=RTX A60002026.01 | 57.3 | — | — | |
| functionary-mediumPlatform=RTX A60002026.01 | 53.2 | — | — | |
| qwen2.5:3bPlatform=RTX A60002026.01 | 13.9 | — | — | |
| functionary-smallPlatform=RTX A60002026.01 | 7.5 | — | — |