Qwen 1.5 110B Chat: benchmark results
Chat-tuned variant of Alibaba's 110B Qwen1.5 flagship (April 2024), roughly on par with Llama-3-70B-Instruct in chat evals. Provider: Alibaba. Released 2024-04-25. Access: Open.
Aggregate Rating 1528 ± 13, rank #691 of 1656 rated models, from 222 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenCompass Language - Intention Recognition - Chinese (CompassBench 2405) | 78.7 | Score (%) | 100 |
| Open LLM Leaderboard v1 - MMLU | 78.04 | Accuracy (%) (5-shot) | 99.1 |
| OpenCompass Subjective - Creation (CompassBench 2404) | 57.5 | Score (%) | 98 |
| OpenCompass Math - High School - Chinese (CompassBench 2405) | 39.5 | Score (%) | 97.4 |
| OpenCompass Subjective - Conversation - English (CompassBench 2405) | 57.4 | Score (%) | 97.4 |
| OpenCompass Language - Dialogue - English (CompassBench 2407) | 57.6 | Score (%) | 97.1 |
| OpenCompass Knowledge - Social Science - Chinese (CompassBench 2405) | 72.2 | Score (%) | 94.7 |
| OpenCompass LLM - Knowledge (CompassBench 2405) | 71.4 | Score (%) | 94.7 |
| OpenCompass LLM - Language - Chinese (CompassBench 2405) | 63.6 | Score (%) | 94.7 |
| OpenCompass Language - Traditional Cultural Understanding - Chinese (CompassBench 2405) | 70.7 | Score (%) | 94.7 |
| OpenCompass Subjective - Conversation (CompassBench 2405) | 59.4 | Score (%) | 94.7 |
| OpenCompass Subjective - Creation - Chinese (CompassBench 2405) | 55.9 | Score (%) | 94.7 |
Interactive version: theaggregate.ai/model?slug=qwen-1-5-110b-chat · How It Works · Data refreshed daily, snapshot 2026-10-20.