MiniCPM5-2B: benchmark results
Provider: OpenBMB. Access: Open.
Unified ELO 1583 ± 36, rank #498 of 1629 rated models, from 59 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA-Omniscience Hallucination Rate | 21.91 | Hallucination Rate (%) | 97.7 |
| AA-Omniscience Index - Law | -5.6 | Omniscience Index | 76.8 |
| AA-Omniscience Index - Software Engineering (SWE) - Dart | 4 | Omniscience Index | 74.8 |
| AA-Omniscience Index - Humanities & Social Sciences | -6.6 | Omniscience Index | 72.6 |
| AA-Omniscience Index - Business | -8.6 | Omniscience Index | 72.3 |
| AA-Omniscience Index - Software Engineering (SWE) - Julia | -12 | Omniscience Index | 70.9 |
| AA-Omniscience Index - Science, Engineering & Mathematics | -7.5 | Omniscience Index | 66.5 |
| AA-Omniscience Index - Software Engineering (SWE) - Java | -26 | Omniscience Index | 66.1 |
| AA Omniscience | -11.65 | Score | 65.3 |
| AA-Omniscience Index - Health | -20.6 | Omniscience Index | 64 |
| AA-Omniscience Index - Software Engineering (SWE) - R | -28 | Omniscience Index | 61.9 |
| AA-Omniscience Index - Software Engineering (SWE) - Go | -16 | Omniscience Index | 61.4 |
Interactive version: theaggregate.ai/model?slug=minicpm5-2b · How It Works · Data refreshed daily, snapshot 2026-10-07.