OmniVerifier-7B: benchmark results
Provider: Other. Access: Open.
Aggregate Rating 1599 ± 48, rank #1024 of 10916 rated models, from 47 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| ViVerBench Release Rule-based - Robot. | 0.73 | Verification score (0–1) | 100 |
| ViVerBench Release Rule-based - LaTeX | 0.91 | Verification score (0–1) | 90 |
| ViVerBench Release Model-based - Maze | 0.5 | Verification score (0–1) | 80 |
| 3DVerifier - Fine Geometry | 51.7 | Balanced accuracy (%) | 72.2 |
| ViVerBench Release Rule-based - Maze | 0.56 | Verification score (0–1) | 70 |
| ViVerBench Release Rule-based - Obj. | 0.73 | Verification score (0–1) | 70 |
| ViVerBench Release Rule-based - Spat. | 0.74 | Verification score (0–1) | 65 |
| ViVerBench Release Rule-based - D-Phy | 0.62 | Verification score (0–1) | 60 |
| ViVerBench Release Model-based - D-Phy | 0.38 | Verification score (0–1) | 50 |
| ViVerBench Release Model-based - Spat. | 0.72 | Verification score (0–1) | 50 |
| ViVerBench Release Model-based - Count | 0.46 | Verification score (0–1) | 45 |
| ViVerBench Release Model-based - Obj. | 0.69 | Verification score (0–1) | 45 |
Interactive version: theaggregate.ai/model?slug=omniverifier-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.