Comprehensive Data Intelligence on Aggregated Benchmarks (DABench, DSBench, TableBench, DABStep-R)
73.68Overall ScoreVeriGraph-8B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| VeriGraph-8BCategory=Specialized data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✓2026.06 | 73.68 | 87.61 | |
| Claude-4.5-OpusCategory=ReAct data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 73.22 | 73.57 | |
| GPT-5.4Category=ReAct data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 72.01 | 78.52 | |
| Claude-4.5-SonnetCategory=ReAct data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 71.17 | 74.16 | |
| Qwen3-32BCategory=ReAct data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 70.98 | 71.95 | |
| DeepAnalyzeCategory=Specialized data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 66.54 | 61.35 | |
| GPT-5.2Category=ReAct data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 64.98 | 74.17 | |
| Qwen3-30B-A3BCategory=ReAct data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 64.36 | 64.08 | |
| QwQ-32BCategory=ReAct data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 63.43 | 69.89 | |
| Qwen3-8BCategory=ReAct data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 62.35 | 59.52 | |
| Claude-4.5-OpusCategory=Direct inference, Traceability (Comp.)=✗, Traceability (Deriv.)=✗2026.06 | 58.85 | — | |
| Claude-4.5-SonnetCategory=Direct inference, Traceability (Comp.)=✗, Traceability (Deriv.)=✗2026.06 | 56.27 | — | |
| Gemini-2.5-ProCategory=Direct inference, Traceability (Comp.)=✗, Traceability (Deriv.)=✗2026.06 | 55.86 | — | |
| DataMindCategory=Specialized data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 55.06 | 71.64 | |
| Qwen3-32BCategory=Direct inference, Traceability (Comp.)=✗, Traceability (Deriv.)=✗2026.06 | 53.8 | — | |
| Qwen3-30B-A3BCategory=Direct inference, Traceability (Comp.)=✗, Traceability (Deriv.)=✗2026.06 | 53.18 | — | |
| Qwen3-Coder-30BCategory=ReAct data agents, Traceability (Comp.)=✓, Traceability (Deriv.)=✗2026.06 | 51.25 | 73.86 | |
| GPT-5.2Category=Direct inference, Traceability (Comp.)=✗, Traceability (Deriv.)=✗2026.06 | 48.75 | — |