List your company

Financial workflow task execution on RealFin benchmark

68AccuracyBA-Inc
33.6842.5951.560.41Apr 18, 2026Apr 26, 2026May 4, 2026May 12, 2026May 20, 2026May 28, 2026Jun 6, 2026

Evaluation results

MethodLinks
BA-IncModel=deepseek-v4-pro2026.06681.591301.721.74
GAModel=Claude Sonnet 4.62026.0465102121145.7
GAModel=Claude Sonnet 4.62026.0665102121145.7
BA-IncModel=deepseek-v4-flash2026.06651.891262.023.96
BA-FullModel=deepseek-v4-pro2026.06653.383233.77.02
Claude CodeModel=Claude Opus 4.62026.0460290173071.95
CodexModel=GPT-5.42026.0460838548920.67
Claude CodeModel=Claude Opus 4.62026.0660290173071.95
CodexModel=GPT-5.42026.0660838548920.67
GAModel=deepseek-v4-pro2026.06603.43173.726.46
Claude CodeModel=Claude Sonnet 4.62026.0455226122382.31
Claude CodeModel=Claude Sonnet 4.62026.0655226122382.31
BA-FullModel=deepseek-v4-flash2026.06522.92443.156.67
GAModel=deepseek-v4-flash2026.06453.982624.244.24
OpenClawModel=Claude Sonnet 4.62026.043524922511.39
OpenClawModel=Claude Sonnet 4.62026.063524922511.39