MiniCPM-o-4.5: benchmark results
Provider: OpenBMB. Released 2026-02-03. Access: Open.
Unified ELO 1606 ± 14, rank #432 of 1605 rated models, from 80 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| DuplexSpeechBench-IFEval (PAS) | 66.2 | Persona Adherence Score (0-100; GPT-4o-judged persona-consis | 100 |
| KMMAU - Topic Summary | 98 | Accuracy (%; four-choice question on the topic of the audio; | 100 |
| MOV-Bench - Hypothetical Reasoning | 40 | Accuracy (%; 60 hypothetical-reasoning questions; direct inf | 100 |
| VideoFDB - Perception | 3.4 | gpt-4o judge score (0-5; the 105 held-out test clips of the | 100 |
| VideoFDB - Perception (Audio Only) | 3.44 | gpt-4o judge score (0-5; the 105 held-out test clips of the | 100 |
| VideoFDB - Perception (Audio Only) - Conversational Flow | 3.76 | gpt-4o judge score (0-5; the 105 held-out test clips of the | 100 |
| VideoFDB - Perception (Audio Only) - Fluency | 3.45 | gpt-4o judge score (0-5; the 105 held-out test clips of the | 100 |
| VideoFDB - Perception - Conversational Flow | 3.54 | gpt-4o judge score (0-5; the 105 held-out test clips of the | 100 |
| VideoFDB - Perception - Semantic Grounding | 3.63 | gpt-4o judge score (0-5; the 105 held-out test clips of the | 100 |
| MathVista | 82.2 | Accuracy (%) | 98.1 |
| HallusionBench | 65 | Overall Score | 97.7 |
| MMStar | 74.5 | Accuracy (%) | 97 |
Interactive version: theaggregate.ai/model?slug=minicpm-o-4-5 · How It Works · Data refreshed daily, snapshot 2026-09-26.