Qwen2.5-7B-Instruct-abliterated: benchmark results
huihui-ai's original abliterated Qwen2.5 7B Instruct with refusals removed, first in a series later revised as v2 and v3. Provider: Alibaba. Released 2024-09-19. Access: Open.
Unified ELO 1495 ± 1, rank #713 of 1392 rated models, from 130 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - MATH Level 5 | 45.77 | Score | 93.9 |
| Open Korean LLM Leaderboard | 45.22 | Average Score (%) | 91.3 |
| Open LLM Leaderboard - IFEval | 75.46 | Score | 90.9 |
| Open Arabic LLM - Madinah QA Arabic Language (Grammar) | 60.82 | Accuracy (%) | 85.8 |
| Open Arabic LLM - Madinah QA Average | 63.01 | Accuracy (%) | 79.6 |
| Open LLM Leaderboard - MMLU-Pro | 35.33 | Score | 76 |
| Open Arabic LLM - Arabic MMLU Arabic Language (Grammar) | 44.11 | Accuracy (%) | 75 |
| Open Arabic LLM - Aratrust Illegal | 90.57 | Accuracy (%) | 73.9 |
| Open LLM Leaderboard - GPQA | 8.72 | Score | 71.9 |
| Open Arabic LLM - Arabic MMLU HT College Chemistry | 43 | Accuracy (%) | 71.3 |
| Open Arabic LLM - Alghafa Multiple Choice Sentiment Task | 39.88 | Accuracy (%) | 66.4 |
| Open Arabic LLM - Aratrust Privacy | 92.98 | Accuracy (%) | 64.9 |
Interactive version: theaggregate.ai/model?slug=qwen2-5-7b-instruct-abliterated · How It Works · Data refreshed daily, snapshot 2026-09-05.