GAOKAO-Eval 2024
Models take Chinese national college entrance examinations. High school teachers grade the open responses.
- Made by
- OpenCompass and collaborators
- Released
- June 2024
- Kind
- Exam questions
- Level
- School
- Scored
- Graded by people
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Capability
- Subject
- Mathematics, Physical sciences, Chemical sciences, Languages and literature, Biological sciences
- Language
- Chinese
- Interaction
- Single turn
- Items from
- Exams and contests
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Biology and medicine, Chemistry and materials
- Strongest model tested
- GPT-4o May 2024 #162 on the capability index
- Last model added
- June 2024
- Scores refreshed
- November 20th, 2026
Qwen2-72B leads with 476.5 among models released by June 2024 Qwen2-72B leads with 476.5 among models released by June 2024
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. The authors rank each paper by science total. The headline averages the New Curriculum and National A science totals. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.
The newest model with a score here came out in June 2024, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1
Qwen2-72B 476.5
- 2 GPT-4o May 2024 476.3
- 3 InternLM2-WQX-20B + InternLM2-WQX-VL-20B 465.3
- 4 Qwen2-72B + Qwen-VL-Chat 428.0
- 5 GLM-4-9B-Chat + GLM-4V-9B 380.3
- 6
Mixtral 8x22B 372.0
- 7 Yi-1.5-34B-Chat + Yi-VL-34B 367.8
- 8 Qwen2-57B-A14B-Instruct + Qwen-VL-Chat 367.0
Source: Official GAOKAO-Eval 2024 leaderboard
- 1
Qwen2-72B 476.5
- 2 GPT-4o May 2024 476.3
- 3
Mixtral 8x22B 372.0
Source: Official GAOKAO-Eval 2024 leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 476.5 | Jun 2024 | Official leaderboard | |
| 2 | GPT-4o May 2024 | 476.3 | May 2024 | Official leaderboard | |
| 3 | InternLM2-WQX-20B + InternLM2-WQX-VL-20B | 465.3 | Official leaderboard | ||
| 4 | Qwen2-72B + Qwen-VL-Chat | 428.0 | Official leaderboard | ||
| 5 | GLM-4-9B-Chat + GLM-4V-9B | 380.3 | Official leaderboard | ||
| 6 | | 372.0 | Apr 2024 | Official leaderboard | |
| 7 | Yi-1.5-34B-Chat + Yi-VL-34B | 367.8 | Official leaderboard | ||
| 8 | Qwen2-57B-A14B-Instruct + Qwen-VL-Chat | 367.0 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 476.5 | Jun 2024 | Official leaderboard | |
| 2 | GPT-4o May 2024 | 476.3 | May 2024 | Official leaderboard | |
| 3 | | 372.0 | Apr 2024 | Official leaderboard |
Rankings by part
We read 22 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as New Curriculum Chinese, New Curriculum Mathematics and New Curriculum English. Qwen2-72B leads 13 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
New Curriculum Chinese
3 models we track have a score here.
- 1
Qwen2-72B 124.0
- 2 GPT-4o May 2024 111.5
- 3
Mixtral 8x22B 77.5
New Curriculum Mathematics
3 models we track have a score here.
- 1 GPT-4o May 2024 ↑1 73.0
- 2
Qwen2-72B ↓1 68.0
- 3
Mixtral 8x22B 21.0
New Curriculum English
3 models we track have a score here.
- 1 GPT-4o May 2024 ↑1 141.5
- 2
Qwen2-72B ↓1 139.0
- 3
Mixtral 8x22B 116.5
New Curriculum Physics
3 models we track have a score here.
- 1
Qwen2-72B 42.0
- 2 GPT-4o May 2024 36.0
- 3
Mixtral 8x22B 25.0
New Curriculum Chemistry
3 models we track have a score here.
- 1
Qwen2-72B 44.0
- 2 GPT-4o May 2024 40.0
- 3
Mixtral 8x22B 35.0
New Curriculum Biology
3 models we track have a score here.
- 1 GPT-4o May 2024 ↑1 65.0
- 2
Qwen2-72B ↓1 48.0
- 3
Mixtral 8x22B 46.0
New Curriculum History
3 models we track have a score here.
- 1 GPT-4o May 2024 ↑1 88.0
- 2
Qwen2-72B ↓1 85.0
- 3
Mixtral 8x22B 54.0
New Curriculum Geography
3 models we track have a score here.
- 1
Qwen2-72B 70.0
- 2 GPT-4o May 2024 59.0
- 3
Mixtral 8x22B 56.0
New Curriculum Politics
3 models we track have a score here.
- 1
Qwen2-72B 60.0
- 2 GPT-4o May 2024 58.0
- 3
Mixtral 8x22B 38.0
New Curriculum Science total
3 models we track have a score here.
- 1 GPT-4o May 2024 ↑1 467.0
- 2
Qwen2-72B ↓1 465.0
- 3
Mixtral 8x22B 321.0
New Curriculum Humanities total
3 models we track have a score here.
- 1
Qwen2-72B 546.0
- 2 GPT-4o May 2024 531.0
- 3
Mixtral 8x22B 363.0
National A Chinese
3 models we track have a score here.
- 1
Qwen2-72B 128.0
- 2 GPT-4o May 2024 122.0
- 3
Mixtral 8x22B 92.0
National A English
3 models we track have a score here.
- 1 GPT-4o May 2024 ↑1 142.5
- 2
Mixtral 8x22B ↑1 142.0
- 3
Qwen2-72B ↓2 141.0
National A Physics
3 models we track have a score here.
- 1
Mixtral 8x22B ↑2 38.0
- 2
Qwen2-72B ↓1 32.0
- 3 GPT-4o May 2024 ↓1 31.0
National A Chemistry
3 models we track have a score here.
- 1
Qwen2-72B 48.0
- 2
Mixtral 8x22B ↑1 39.0
- 3 GPT-4o May 2024 ↓1 34.0
National A Biology
3 models we track have a score here.
- 1 GPT-4o May 2024 ↑1 72.0
- 2
Mixtral 8x22B ↑1 54.0
- 3
Qwen2-72B ↓2 50.0
National A History
3 models we track have a score here.
- 1 GPT-4o May 2024 ↑1 82.0
- 2
Mixtral 8x22B ↑1 74.0
- 3
Qwen2-72B ↓2 71.0
National A Geography
3 models we track have a score here.
- 1
Qwen2-72B 81.0
- 2
Mixtral 8x22B ↑1 74.0
- 3 GPT-4o May 2024 ↓1 66.0
National A Science mathematics
3 models we track have a score here.
- 1
Qwen2-72B 89.0
- 2 GPT-4o May 2024 84.0
- 3
Mixtral 8x22B 58.0
National A Humanities mathematics
3 models we track have a score here.
- 1
Qwen2-72B 95.0
- 2 GPT-4o May 2024 89.0
- 3
Mixtral 8x22B 53.0
National A Science total
3 models we track have a score here.
- 1
Qwen2-72B 488.0
- 2 GPT-4o May 2024 485.5
- 3
Mixtral 8x22B 423.0
National A Humanities total (excluding politics)
3 models we track have a score here.
- 1
Qwen2-72B 516.0
- 2 GPT-4o May 2024 501.5
- 3
Mixtral 8x22B 435.0
2 models have held number one on the official leaderboard since its release in June 2024
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official GAOKAO-Eval 2024 leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official GAOKAO-Eval 2024 leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Knowledge
- MMLU-Pro In use Harder 10-option version of MMLU with reasoning-focused questions across 14 subjects.
Intern-S2-397B 89.8%
- MMLU In use 57-subject multiple-choice exam questions from elementary to professional level. Scored by accuracy.
Seed1.8 92.3%
- SuperGPQA In use 26,529 graduate-level multiple-choice questions across 285 disciplines. Accuracy.
Qwen 3.6 Max (Preview) 73.9%
- WideSearch In use 200 bilingual broad info-seeking tasks: agent collects many atomic facts into a table. Scored by success rate and row/item F1.
Tencent Hy4 preview 83.9%
- BBEH micro average In use Models solve harder replacements for each of the 23 BIG-Bench Hard tasks, 4,520 examples in total. The score is the share of all examples answered correctly.
Gemma 4 31B IT 74.4%
- HLE-Diamond In use Models answer 1,000 cleaned expert questions from Humanity's Last Exam, 500 on reasoning and 500 on knowledge, closed book. Graded on exact answers. Claude Fable 5.2 77.9%
Data audit
8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 2 years ago, in June 2024, for Qwen2-72B. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (8 entries).
- Pass Models scored 3 models we track, plus 5 other entries on its leaderboard.
- Warning Newest score The newest came 2 years ago, in June 2024, for Qwen2-72B.
- Note Frontier models None of the top 10 has a score. The strongest tested is GPT-4o (#162).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on github.com: 3 rows for models we track and 5 other entries at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.