abab6.5-chat: benchmark results
Provider: MiniMax. Access: API.
Aggregate Rating 1517 ± 27, rank #2098 of 10916 rated models, from 55 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenCompass Reasoning - Common Sense - English (CompassBench 2407) | 49.5 | Score (%) | 94.1 |
| OpenCompass LLM - Reasoning (CompassBench 2407) | 47 | Score (%) | 79.4 |
| OpenCompass LLM - Reasoning - English (CompassBench 2407) | 48 | Score (%) | 79.4 |
| OpenCompass Reasoning - Science and Engineering - Chinese (CompassBench 2407) | 48.6 | Score (%) | 79.4 |
| OpenCompass Reasoning - Social - English (CompassBench 2407) | 49.3 | Score (%) | 79.4 |
| OpenCompass Reasoning - Science and Engineering - English (CompassBench 2407) | 50 | Score (%) | 76.5 |
| OpenCompass Code - HumanEval+ - English (CompassBench 2407) | 72.6 | Score (%) | 70.6 |
| OpenCompass LLM - Reasoning - Chinese (CompassBench 2407) | 45.9 | Score (%) | 70.6 |
| OpenCompass Reasoning - Common Sense - Chinese (CompassBench 2407) | 44.3 | Score (%) | 70.6 |
| OpenCompass Language - NLP - English (CompassBench 2407) | 46.8 | Score (%) | 69.1 |
| OpenCompass Code - LCBench - Chinese (CompassBench 2407) | 38.3 | Score (%) | 67.6 |
| OpenCompass LLM - Code - English (CompassBench 2407) | 52.1 | Score (%) | 67.6 |
Interactive version: theaggregate.ai/model?slug=abab6-5-chat · How It Works · Data refreshed daily, snapshot 2026-11-08.