Qwen in-house Chinese automatic evaluation
Qwen2-72B 69.6% Models answer Qwen in-house Chinese datasets on knowledge, comprehension, coding, math and related skills. Skills and weights change between reports.
In-house Qwen evaluation. The Qwen2 report gives an average, the Qwen2.5 report gives per-skill scores only.
- Made by
- Qwen Team, Alibaba Group
- Released
- July 2024
- Kind
- Chat and instructions in many languages
- Level
- Everyday
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- Kept private
- Strongest model tested
- Qwen2-72B #179 on the capability index
- Last in a system card
- December 2024
- Last model added
- December 2024
Qwen2-72B leads with 69.6%
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
- 1 Qwen2-72B 69.6%
Source: Labs’ model cards
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Qwen2-72B | 69.6% | Jul 2024 | Lab model card |
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Labs’ model cards
Labs report it only in system cards
We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark. Some system cards do, below.
6 system cards report it, from 1 lab
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Dec 19, 2024 | Qwen2.5-7B technical report | In-house Chinese automatic evaluation, per-skill scores only, no overall: IF 26.64, Knowledge 65.77, Comprehension 67.55, Coding 39.56, Math 61.06, Reasoning 49.70 (Table 12) | Capabilities | |
| Dec 19, 2024 | Qwen2.5-72B technical report | In-house Chinese automatic evaluation, per-skill scores only, no overall: IF 37.22, Knowledge 75.86, Comprehension 78.85, Coding 56.71, Math 68.39, Reasoning 63.02 (Table 12) | Capabilities | |
| Dec 19, 2024 | Qwen2.5-32B technical report | In-house Chinese automatic evaluation, per-skill scores only, no overall: IF 32.64, Knowledge 74.70, Comprehension 79.46, Coding 54.45, Math 67.86, Reasoning 60.19 (Table 12) | Capabilities | |
| Dec 19, 2024 | Qwen2.5-14B technical report | In-house Chinese automatic evaluation, per-skill scores only, no overall: IF 26.87, Knowledge 70.28, Comprehension 76.96, Coding 49.78, Math 67.01, Reasoning 56.41 (Table 12) | Capabilities | |
| Dec 19, 2024 | Qwen-Turbo technical report | In-house Chinese automatic evaluation, per-skill scores only, no overall: IF 32.94, Knowledge 72.93, Comprehension 74.37, Coding 51.92, Math 66.08, Reasoning 53.30 (Table 12) | Capabilities | |
| Jul 15, 2024 | Qwen2-72B technical report | 69.6 | In-house Chinese automatic evaluation Avg.: Knowledge 76.19, Exam 75.65, Comprehension 74.72, Coding 49.53, Math 70.80, Reasoning 70.59 (Table 10) | Capabilities |
More in Multilingual
- MMMLU In use MMLU professionally translated into 14 languages. Multiple-choice accuracy. Claude Mythos Preview 92.7%
- WMT24++ In use English-to-55-language translation of WMT24 test data. Scored with automatic metrics such as MetricX or xCOMET.
K-EXAONE 90.5% - MMLU-ProX In use MMLU-Pro translated into up to 29 languages with expert review. Multiple-choice accuracy per language. Qwen3.7-Max 87%
- CLIcK In use Models answer 1,995 multiple-choice questions on Korean language and culture. Questions come from official Korean exams and textbooks in eleven categories.
Solar Open 2 90.7% - Global MMLU Lite 3.0 In use Models answer 400 human-translated MMLU questions per language in 23 languages. Version 3.0 adds Czech, Hungarian, Oriya, Slovak and Tajik and updates Italian. Accuracy.
Nemotron 3 Ultra 550B-A55B Base 90.1% - MMMB In use Models answer multiple-choice questions about images in six languages across 15 categories. The set has about 12,000 questions. LFM2.5-VL-3B 83%