Benchmark Almanac

SuperCLUE-Hard

o1 76.7 leads since December 2024
Not in model cards

Models follow precise instructions, solve reasoning problems and plan tool calls in Chinese. The report gives separate scores for each ability.

The annual report ranks ChatGPT, Gemini, Claude, Step and DeepSeek models in the December 2024 evaluation.

Made by
SuperCLUE
Released
December 2024
Kind
Reasoning in many languages
Level
University
Scored
Graded by a model
Answers
Open-ended
Measures
Capability
Language
Chinese
In and out
Text → Text, Actions
Interaction
Agentic
Environment
APIs and tools
Refresh
One fixed set
Run by
A host
In views
Agents
Strongest model tested
o1 #118 on the capability index
Last model added
March 2025
Scores refreshed
November 21st, 2026
40
Entries in its paper, 13 models we track

o1 leads with 76.7 among models released by March 2025 o1 leads with 76.7 among models released by March 2025

Best score per entry from the benchmark's paper, read on November 21st, 2026. A paper’s table doesn’t add newer models. 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 March 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 o1 76.7
  2. 2 o1-preview 63.6
  3. 3 ChatGPT-4o-latest 57.8
  4. 4 Gemini 2.0 Flash Feb 2025 55.5
  5. 5 DeepSeek-V3 Mar 2025 54.8
  6. 6 Claude 3.5 Sonnet(20241022) 54.6
  7. 7 360zhinao2-o1 51.5
  8. 8 Doubao-pro-32k-241215 51.5
  9. 9 Qwen-max-latest 51.5
  10. 10 SenseChat 5.5-latest 51.5
  11. 11 Step-2-16k 50.0
  12. 12 Qwen2.5-72B 49.7
  13. 13 Grok-2 Dec 2024 49.2
  14. 14 GLM-4-Plus 48.6
  15. 15 NebulaCoder-V5 48.6
  16. 16 Baichuan4 45.6
  17. 17 ERNIE-4.0-Turbo-8K-Latest 45.6
  18. 18 DeepSeek-V2.5 Sep 2024 45.3
  19. 19 kimi 44.5
  20. 20 Sky-Chat-3.0 44.5
  21. 21 MiniMax-abab7-preview 43.3
  22. 22 TeleChat2-Large 43.3
  23. 23 GPT-4o mini 42.8
  24. 24 Llama 3.3 70B Instruct 38.8
  25. 25 Hunyuan-Turbo 38.6
  26. 26 TeleChat2-35B 38.6
  27. 27 Qwen2.5-7B-Instruct 35.7
  28. 28 GLM-4-9B-Chat 31.6
  29. 29 QwQ-32B-Preview 26.6
  30. 30 Gemma-2-9b-it 22.7
  31. 31 Llama-3.1-8B-Instruct 20.9
  32. 32 Yi-1.5-34B-Chat-16K 20.6
  33. 33 Yi-1.5-9B-Chat-16K 20.6
  34. 34 讯飞星火V4.0 20.6
  35. 35 Qwen2.5-3B-Instruct 18.6
  36. 36 360Zhinao2-7B-Chat-4K 17.5
  37. 37 Phi-3.5-Mini-Instruct 14.0
  38. 38 MiniCPM3-4B 13.7
  39. 39 Mistral-7B-Instruct-v0.3 11.8
  40. 40 Gemma 2 2B IT 11.8

Source: The SuperCLUE-Hard paper

#EntryScoreSettingDateSource
1 o1 76.7 Dec 2024 Official leaderboard
2 o1-preview 63.6 Sep 2024 Official leaderboard
3 ChatGPT-4o-latest 57.8 Official leaderboard
4 Gemini 2.0 Flash Feb 2025 55.5 Feb 2025 Official leaderboard
5 DeepSeek-V3 Mar 2025 54.8 Mar 2025 Official leaderboard
6 Claude 3.5 Sonnet(20241022) 54.6 Official leaderboard
7 360zhinao2-o1 51.5 Official leaderboard
8 Doubao-pro-32k-241215 51.5 Official leaderboard
9 Qwen-max-latest 51.5 Official leaderboard
10 SenseChat 5.5-latest 51.5 Official leaderboard
11 Step-2-16k 50.0 Official leaderboard
12 Qwen2.5-72B 49.7 Sep 2024 Official leaderboard
13 Grok-2 Dec 2024 49.2 Dec 2024 Official leaderboard
14 GLM-4-Plus 48.6 Official leaderboard
15 NebulaCoder-V5 48.6 Official leaderboard
16 Baichuan4 45.6 Official leaderboard
17 ERNIE-4.0-Turbo-8K-Latest 45.6 Official leaderboard
18 DeepSeek-V2.5 Sep 2024 45.3 Sep 2024 Official leaderboard
19 kimi 44.5 Official leaderboard
20 Sky-Chat-3.0 44.5 Official leaderboard
21 MiniMax-abab7-preview 43.3 Official leaderboard
22 TeleChat2-Large 43.3 Official leaderboard
23 GPT-4o mini 42.8 Jul 2024 Official leaderboard
24 Llama 3.3 70B Instruct 38.8 Dec 2024 Official leaderboard
25 Hunyuan-Turbo 38.6 Official leaderboard
26 TeleChat2-35B 38.6 Official leaderboard
27 Qwen2.5-7B-Instruct 35.7 Official leaderboard
28 GLM-4-9B-Chat 31.6 Jun 2024 Official leaderboard
29 QwQ-32B-Preview 26.6 Nov 2024 Official leaderboard
30 Gemma-2-9b-it 22.7 Official leaderboard
31 Llama-3.1-8B-Instruct 20.9 Official leaderboard
32 Yi-1.5-34B-Chat-16K 20.6 Official leaderboard
33 Yi-1.5-9B-Chat-16K 20.6 Official leaderboard
34 讯飞星火V4.0 20.6 Official leaderboard
35 Qwen2.5-3B-Instruct 18.6 Official leaderboard
36 360Zhinao2-7B-Chat-4K 17.5 Official leaderboard
37 Phi-3.5-Mini-Instruct 14.0 Official leaderboard
38 MiniCPM3-4B 13.7 Sep 2024 Official leaderboard
39 Mistral-7B-Instruct-v0.3 11.8 Official leaderboard
40 Gemma 2 2B IT 11.8 Jul 2024 Official leaderboard
#ModelScoreSettingDateSource
1 o1 76.7 Dec 2024 Official leaderboard
2 o1-preview 63.6 Sep 2024 Official leaderboard
3 Gemini 2.0 Flash Feb 2025 55.5 Feb 2025 Official leaderboard
4 DeepSeek-V3 Mar 2025 54.8 Mar 2025 Official leaderboard
5 Qwen2.5-72B 49.7 Sep 2024 Official leaderboard
6 Grok-2 Dec 2024 49.2 Dec 2024 Official leaderboard
7 DeepSeek-V2.5 Sep 2024 45.3 Sep 2024 Official leaderboard
8 GPT-4o mini 42.8 Jul 2024 Official leaderboard
9 Llama 3.3 70B Instruct 38.8 Dec 2024 Official leaderboard
10 GLM-4-9B-Chat 31.6 Jun 2024 Official leaderboard
11 QwQ-32B-Preview 26.6 Nov 2024 Official leaderboard
12 MiniCPM3-4B 13.7 Sep 2024 Official leaderboard
13 Gemma 2 2B IT 11.8 Jul 2024 Official leaderboard

2 models have held number one in the benchmark’s paper since its release in December 2024

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI
Benchmark paper Not tested since Mar 2025
o1 76.7

Source: The SuperCLUE-Hard paper

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
In the benchmark’s paper
0.00 20.0 40.0 60.0 80.0 2024 2025 0.00 20.0 40.0 60.0 80.0 2024 2025

Source: Official SuperCLUE-Hard 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 Multilingual

All 545 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. Claude Opus 5 95.7%
  • 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%

Data audit

7 of 9 checks pass, 2 warnings. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark

Catalog record 1 warning 1 pass
  • Warning Task access We don’t know whether the tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the benchmark’s paper (40 entries).
  • Pass Models scored 13 models we track, plus 27 other entries on its leaderboard.
  • Warning Newest score The newest came 20 months ago, in March 2025, for DeepSeek-V3.
  • Note Frontier models None of the top 10 has a score. The strongest tested is o1 (#118).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on cluebenchmarks.com: 13 rows for models we track and 27 other entries at the last read.
  • Pass Last read November 21st, 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.

How we check