Finance & Accounting workflows on FINCH 172 workflows
95AccuracyClaude Opus 4.6
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Claude Opus 4.6Technique=BRTR2026.03 | 95 | 37 | 358 | 4.8 | |
| GPT-5.2Technique=BRTR2026.03 | 95 | 33 | 223 | 4.3 | |
| Claude Sonnet 4.5Technique=BRTR2026.03 | 87 | 25 | 305 | 2.7 | |
| Gemini 3 ProTechnique=BRTR2026.03 | 73 | 28 | 380 | 7.9 | |
| Claude Opus 4.6Technique=Naïve2026.03 | 63 | — | — | — | |
| GPT-5.2Technique=Naïve2026.03 | 46 | — | — | — | |
| Gemini 3 ProTechnique=Naïve2026.03 | 27 | — | — | — | |
| Claude Sonnet 4.5Technique=Naïve2026.03 | 20 | — | — | — | |
| LLaMA 4 Maverick 17BTechnique=BRTR2026.03 | 12 | 8 | 113 | 0.5 | |
| LLaMA 4 Maverick 17BTechnique=Naïve2026.03 | 1 | — | — | — |