Qwen in-house English automatic evaluation
Qwen2-72B 72.9% Only in system cards Models answer Qwen in-house English 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 Jul 2024 Other Tasks kept private Last in a system card Dec 2024
Last model added Dec 2024
- Level
- Everyday Most adults, without training
- Scored
- Graded by a model A language model scores each response, usually against a rubric.
- Answers
- Open-ended Free-form text, documents, designs or media judged as a whole.
1
Model scored in lab cards
6
System cards report it, from 1 lab
Qwen2-72B leads with 72.9%
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 72.9%
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Qwen2-72B | 72.9% | Jul 2024 | Model card |
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Reported by the lab in its model card
Labs report it only in system cards
We read the headline results of 479 launch posts from 51 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 |
|---|---|---|---|---|
| 19 Dec 2024 | Qwen2.5-7B technical report | In-house English automatic evaluation, per-skill scores only, no overall: IF 70.01, Knowledge 52.74, Comprehension 62.69, Coding 48.41, Math 56.93, Reasoning 54.69 (Table 11) | Capabilities | |
| 19 Dec 2024 | Qwen2.5-72B technical report | In-house English automatic evaluation, per-skill scores only, no overall: IF 82.65, Knowledge 66.09, Comprehension 74.43, Coding 60.41, Math 59.73, Reasoning 65.90 (Table 11) | Capabilities | |
| 19 Dec 2024 | Qwen2.5-32B technical report | In-house English automatic evaluation, per-skill scores only, no overall: IF 76.79, Knowledge 64.08, Comprehension 71.28, Coding 58.90, Math 60.97, Reasoning 65.49 (Table 11) | Capabilities | |
| 19 Dec 2024 | Qwen2.5-14B technical report | In-house English automatic evaluation, per-skill scores only, no overall: IF 74.17, Knowledge 59.78, Comprehension 69.11, Coding 52.68, Math 59.68, Reasoning 62.51 (Table 11) | Capabilities | |
| 19 Dec 2024 | Qwen-Turbo technical report | In-house English automatic evaluation, per-skill scores only, no overall: IF 72.76, Knowledge 58.56, Comprehension 68.70, Coding 54.48, Math 57.77, Reasoning 61.06 (Table 11) | Capabilities | |
| 15 Jul 2024 | Qwen2-72B technical report | 72.9 | In-house English automatic evaluation Avg.: Knowledge 83.00, Comprehension 73.58, Coding 53.03, Math 82.15 (Table 11) | Capabilities |
More in Other
- Personal Superintelligence Index (Meta) In use Meta's composite of seven personal and knowledge-work evals: Agents' Last Exam, AutomationBench-AA, FocusBench, JobBench, GDP.pdf, VBT and DRACO.
- Artificial Analysis Intelligence Index v5 In use Version 5 of Artificial Analysis's index: nine evaluations in four equal categories, most with private test sets, run three times each in its own agent harness.
Gemini 4 Carbon 59.0 - Anthropic ECI (AECI) Only in system cards Index that combines a basket of capability benchmarks into one capability score to track trends across Anthropic models. Claude Opus 5.5 169.4
- Recognition of model-written user turns (Anthropic) Only in system cards Models write synthetic user turns for real Claude Code transcripts, then judge which version is real. Score is accuracy above chance. Claude Mythos Preview 0.35
- LiveBench Dropped Models answer monthly refreshed questions across math, coding, reasoning, language, instruction following, and data analysis. Scored automatically against objective ground truth. Claude Fable 5.1 85.7%
- PhiBench Dropped Models answer original prompts that Microsoft's Phi team wrote to test a wide range of skills and reasoning.
Phi-4-Reasoning-plus 74.2%