SpaceR-7B: benchmark results
Provider: Other.
Aggregate Rating 1588 ± 35, rank #1138 of 10916 rated models, from 26 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Metric-Bench - Distance | 22.15 | Mean relative accuracy (%; inter-object distance questions; | 85.7 |
| CourtSI-Bench - Ball-Zone | 59.6 | Exact-match accuracy (%) on the 255 ball-to-court-zone relat | 78.3 |
| Metric-Bench - Depth | 52.49 | Mean relative accuracy (%; object-depth questions; 1,340 tem | 71.4 |
| CourtSI-Bench - Ball-Player | 58.6 | Exact-match accuracy (%) on the 297 ball-to-player relation | 63.6 |
| CourtSI-Bench - Player-Player | 59.2 | Exact-match accuracy (%) on the 393 player-to-player relatio | 63.6 |
| CourtSI-Bench - Player-Zone | 40.2 | Exact-match accuracy (%) on the 82 player-to-court-zone rela | 58.3 |
| CourtSI-Bench - Camera-Player | 72.2 | Exact-match accuracy (%) on the 248 camera-to-player relatio | 54.3 |
| GeoReader Spatial Suite - ViewSpatial | 45.5 | Accuracy (%) | 47.4 |
| CourtSI-Bench - Player-Line | 52.3 | Exact-match accuracy (%) on the 495 player-to-court-line rel | 45.5 |
| CourtSI-Bench - Ball Counting | 39.2 | Exact-match accuracy (%) on the 28 ball-counting questions, | 41.7 |
| CourtSI-Bench - Camera-Object Distance | 0.4 | Threshold mean relative accuracy (%), averaged over confiden | 40.9 |
| GeoReader Spatial Suite - SPAR | 37.1 | Accuracy (%) | 36.8 |
Interactive version: theaggregate.ai/model?slug=spacer-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.