Quasar 438B: benchmark results
Provider: Other. Released 2026-09-02.
Unified ELO 1744 ± 35, rank #135 of 1607 rated models, from 61 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA-Omniscience Hallucination Rate | 21.44 | Hallucination Rate (%) | 98 |
| BenchmarkList ECI | 134.57 | Capability Index (ECI) | 84.2 |
| AA-Omniscience Index - Health | -5.9 | Omniscience Index | 82.5 |
| AA-Omniscience Index - Science, Engineering & Mathematics | 5.1 | Omniscience Index | 82.1 |
| AA CritPt | 9.43 | Accuracy (%) | 81.4 |
| Artificial Analysis Intelligence Index | 26.74 | Intelligence Index | 80.8 |
| AA-LCR | 75 | Accuracy (self-reported) | 80.4 |
| AA-Omniscience Index - Business | -4.3 | Omniscience Index | 78.3 |
| AA Long Context Reasoning | 76.33 | Accuracy (%) | 77.4 |
| AA-Omniscience Index - Law | -6.2 | Omniscience Index | 76.8 |
| AA Omniscience | -2.6 | Score | 76.3 |
| AA-Omniscience Index - Software Engineering (SWE) - HTML | 30 | Omniscience Index | 76.1 |
Interactive version: theaggregate.ai/model?slug=quasar-438b · How It Works · Data refreshed daily, snapshot 2026-09-29.