Qwen 2.5 7B Turbo Instruct: benchmark results
Provider: Alibaba. Released 2024-09-19.
Aggregate Rating 1484 ± 20, rank #2610 of 10916 rated models, from 702 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Enkrypt AI - Safety Risk | 17.33 | Risk Score | 93.5 |
| Enkrypt AI - Bias Risk | 50.65 | Risk Score | 91.9 |
| HELM Lite - MATH Level 1 | 83.51 | Equivalent (CoT) (%) | 90 |
| PARSE (Persian Zero-Shot) - Boolean Reasoning Negation | 0.82 | Accuracy (0-1; 600 yes/no reasoning questions with negation; | 90 |
| HELM Lite - MATH Level 1 - Counting and Probability | 89.74 | Equivalent (CoT) (%) | 87.8 |
| HELM Lite - MATH Level 1 - Number Theory | 86.67 | Equivalent (CoT) (%) | 87.2 |
| HELM Lite - MATH Level 1 - Algebra | 96.3 | Equivalent (CoT) (%) | 85.6 |
| HELM AIR-Bench 2024 - #17.2: Suggestive | 70 | Refusal Rate (%) | 83.7 |
| HELM Lite - MATH Level 1 - Intermediate Algebra | 75 | Equivalent (CoT) (%) | 82.8 |
| HELM Lite - MATH Level 1 - Geometry | 68.42 | Equivalent (CoT) (%) | 82.2 |
| HELM MMLU - Business Ethics | 82 | EM | 82.1 |
| HELM Lite - MATH Level 1 - Prealgebra | 93.02 | Equivalent (CoT) (%) | 81.1 |
Interactive version: theaggregate.ai/model?slug=qwen-2-5-7b-turbo-instruct · How It Works · Data refreshed daily, snapshot 2026-11-08.