Beam 1.0: benchmark results
Provider: Reflection. Released 2026-10-16. Access: Open.
Aggregate Rating 1763 ± 33, rank #97 of 1647 rated models, from 77 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Reflection Beam Safety - HPCT - 2026-07-09 - Exact Match | 61 | Accuracy (%) | 100 |
| AA-Omniscience Index - Software Engineering (SWE) - Rust | 64 | Omniscience Index | 85.4 |
| AA-Omniscience Index - Software Engineering (SWE) - PHP | 46 | Omniscience Index | 84.3 |
| AA-Omniscience Index - Software Engineering (SWE) - Swift | 48 | Omniscience Index | 84.2 |
| Artificial Analysis Intelligence Index | 30.33 | Intelligence Index | 82.8 |
| AA CritPt | 13.14 | Accuracy (%) | 82.5 |
| AA-Omniscience Index - Software Engineering (SWE) - HTML | 48 | Omniscience Index | 81.6 |
| AA Humanity's Last Exam | 34.66 | Accuracy (%) | 81.5 |
| AA-Omniscience Index - Software Engineering (SWE) - R | 16 | Omniscience Index | 81.4 |
| AA-Omniscience Hallucination Rate | 47.16 | Hallucination Rate (%) | 81.3 |
| AA-Omniscience Index - Software Engineering (SWE) - C | 55 | Omniscience Index | 81.3 |
| AA-Omniscience Index - Software Engineering (SWE) - Kotlin | 30 | Omniscience Index | 81.1 |
Interactive version: theaggregate.ai/model?slug=beam-1-0 · How It Works · Data refreshed daily, snapshot 2026-10-17.