Benchmark Almanac

OlymMATH

Gemini 2.5 Pro Mar 2025 74.2% leads since March 2025
Not in model cards

Models solve Olympiad-level mathematics problems in English or Chinese. The evaluator checks final answers and reports pass@1 and consistency across repeated samples.

Easy and hard subsets exist in both languages. The released model-output collection is separate from the benchmark questions.

Released
March 2025
Kind
Competition problems with answers
Level
Competition
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Measures
Capability
Subject
Mathematics
Language
English, Chinese
In and out
Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Strongest model tested
Gemini 2.5 Pro Mar 2025 #104 on the capability index
Last model added
April 2025
Scores refreshed
November 20th, 2026

In the authors’ words

The rapid advancement of large reasoning models has saturated existing math benchmarks, underscoring the urgent need for more challenging evaluation frameworks. To address this, we introduce OlymMATH, a rigorously curated, Olympiad-level math benchmark comprising 350 problems, each with parallel English and Chinese versions. OlymMATH is the first benchmark to unify dual evaluation paradigms within a single suite: (1) natural language evaluation through OlymMATH-EASY and OlymMATH-HARD, comprising 200 computational problems with numerical answers for objective rule-based assessment, and (2) formal verification through OlymMATH-LEAN, offering 150 problems formalized in Lean 4 for rigorous process-level evaluation. All problems are manually sourced from printed publications to minimize data contamination, verified by experts, and span four core domains. Extensive experiments reveal the benchmark's significant challenge, and our analysis also uncovers consistent performance gaps between languages and identifies cases where models employ heuristic "guessing" rather than rigorous reasoning. To further support community research, we release 582k+ reasoning trajectories, a visualization tool, and expert solutions at https://github.com/RUCAIBox/OlymMATH.

Abstract of Challenging the Boundaries of Reasoning: An Olympiad-Level Math Benchmark for Large Language Models (arXiv:2503.21380, CC0)
28
Entries in its paper, 13 models we track

Gemini 2.5 Pro leads with 74.2% among models released by April 2025 Gemini 2.5 Pro leads with 74.2% among models released by April 2025

Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. The headline is the plain mean of English easy, English hard, Chinese easy and Chinese hard pass@1. 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 April 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 Gemini 2.5 Pro Mar 2025 74.2%
  2. 2 o3-mini High reasoning effort 61.5%
  3. 3 Qwen3-235B-A22B Thinking 60.6%
  4. 4 Qwen3-30B-A3B Thinking 53.3%
  5. 5 QwQ-32B 48.9%
  6. 6 OpenMath-Nemo. (14B) 48.6%
  7. 7 Skywork-OR1-Pre. (32B) 47.5%
  8. 8 DeepSeek-R1 46.4%
  9. 9 OpenMath-Nemo. (7B) 46.2%
  10. 10 GLM-Z1-32B-0414 43.6%
  11. 11 Light-R1-DS (32B) 43.3%
  12. 12 OpenThinker2 (32B) 40.3%
  13. 13 OpenMath-Nemo. (32B) 40.1%
  14. 14 Qwen3-4B Thinking 38.7%
  15. 15 Light-R1-DS (14B) 33.8%
  16. 16 DeepSeek-R1-Distill-Qwen-32B 32.6%
  17. 17 AceMath-RL (7B) 32.3%
  18. 18 OpenMath-Nemo. (1.5B) 31.4%
  19. 19 Skywork-OR1-Math (7B) 30.7%
  20. 20 DeepSeek-R1-Distill-Qwen-14B 29.6%
  21. 21 OpenThinker2 (7B) 29.4%
  22. 22 Light-R1-DS (7B) 27.6%
  23. 23 Skywork-OR1-Pre. (7B) 26.5%
  24. 24 DeepSeek-R1-Distill-Qwen-7B 23.9%
  25. 25 DeepScaleR-Pre. (1.5B) 10.8%
  26. 26 STILL-3-Pre. (1.5B) 9.00%
  27. 27 DeepSeek-R1-Distill-Qwen-1.5B 6.88%
  28. 28 Qwen3-0.6B Thinking 5.13%

Source: The OlymMATH paper

#EntryScoreSettingDateSource
1 Gemini 2.5 Pro Mar 2025 74.2% Mar 2025 Official leaderboard
2 o3-mini 61.5% High reasoning effort Jan 2025 Official leaderboard
3 Qwen3-235B-A22B 60.6% Thinking Apr 2025 Official leaderboard
4 Qwen3-30B-A3B 53.3% Thinking Apr 2025 Official leaderboard
5 QwQ-32B 48.9% Mar 2025 Official leaderboard
6 OpenMath-Nemo. (14B) 48.6% Official leaderboard
7 Skywork-OR1-Pre. (32B) 47.5% Official leaderboard
8 DeepSeek-R1 46.4% Jan 2025 Official leaderboard
9 OpenMath-Nemo. (7B) 46.2% Official leaderboard
10 GLM-Z1-32B-0414 43.6% Apr 2025 Official leaderboard
11 Light-R1-DS (32B) 43.3% Official leaderboard
12 OpenThinker2 (32B) 40.3% Official leaderboard
13 OpenMath-Nemo. (32B) 40.1% Official leaderboard
14 Qwen3-4B 38.7% Thinking Apr 2025 Official leaderboard
15 Light-R1-DS (14B) 33.8% Official leaderboard
16 DeepSeek-R1-Distill-Qwen-32B 32.6% Jan 2025 Official leaderboard
17 AceMath-RL (7B) 32.3% Official leaderboard
18 OpenMath-Nemo. (1.5B) 31.4% Official leaderboard
19 Skywork-OR1-Math (7B) 30.7% Official leaderboard
20 DeepSeek-R1-Distill-Qwen-14B 29.6% Jan 2025 Official leaderboard
21 OpenThinker2 (7B) 29.4% Official leaderboard
22 Light-R1-DS (7B) 27.6% Official leaderboard
23 Skywork-OR1-Pre. (7B) 26.5% Official leaderboard
24 DeepSeek-R1-Distill-Qwen-7B 23.9% Jan 2025 Official leaderboard
25 DeepScaleR-Pre. (1.5B) 10.8% Official leaderboard
26 STILL-3-Pre. (1.5B) 9.00% Official leaderboard
27 DeepSeek-R1-Distill-Qwen-1.5B 6.88% Jan 2025 Official leaderboard
28 Qwen3-0.6B 5.13% Thinking Apr 2025 Official leaderboard
#ModelScoreSettingDateSource
1 Gemini 2.5 Pro Mar 2025 74.2% Mar 2025 Official leaderboard
2 o3-mini 61.5% High reasoning effort Jan 2025 Official leaderboard
3 Qwen3-235B-A22B 60.6% Thinking Apr 2025 Official leaderboard
4 Qwen3-30B-A3B 53.3% Thinking Apr 2025 Official leaderboard
5 QwQ-32B 48.9% Mar 2025 Official leaderboard
6 DeepSeek-R1 46.4% Jan 2025 Official leaderboard
7 GLM-Z1-32B-0414 43.6% Apr 2025 Official leaderboard
8 Qwen3-4B 38.7% Thinking Apr 2025 Official leaderboard
9 DeepSeek-R1-Distill-Qwen-32B 32.6% Jan 2025 Official leaderboard
10 DeepSeek-R1-Distill-Qwen-14B 29.6% Jan 2025 Official leaderboard
11 DeepSeek-R1-Distill-Qwen-7B 23.9% Jan 2025 Official leaderboard
12 DeepSeek-R1-Distill-Qwen-1.5B 6.88% Jan 2025 Official leaderboard
13 Qwen3-0.6B 5.13% Thinking Apr 2025 Official leaderboard

Rankings by part

We read 20 parts from the benchmark’s paper in 5 groups: english hard, language and difficulty, english easy, chinese hard and chinese easy. Each ranks models on one part of the test. Gemini 2.5 Pro leads 15 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.

Algebra

13 models we track have a score here.

  1. 1 Gemini 2.5 Pro Mar 2025 71.5%
  2. 2 Qwen3-235B-A22B ↑1 48%
  3. 3 Qwen3-30B-A3B ↑1 38.8%
  4. 4 GLM-Z1-32B-0414 ↑3 35%
  5. 5 QwQ-32B 32.9%
  6. 6 DeepSeek-R1 30%
  7. 7 o3-mini ↓5 29.5%
  8. 8 DeepSeek-R1-Distill-Qwen-32B ↑1 22.4%
  9. 9 Qwen3-4B ↓1 18.1%
  10. 10 DeepSeek-R1-Distill-Qwen-14B 16.1%

2 models have held number one in the benchmark’s paper since its release in March 2025

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

OpenAI Google DeepMind
Benchmark paper Not tested since Apr 2025

Source: The OlymMATH 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 Best so far
0% 20% 40% 60% 80% 100% 2025 Near the ceiling (85%) Gemini 2.5 Pro o3-mini DeepSeek-R1 0% 20% 40% 60% 80% 100% 2025 Near the ceiling (85%) Gemini 2.5 Pro o3-mini

Source: Official OlymMATH 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 Math

All 140 in Math
  • AIME 2026 In use Models solve the 30 integer-answer problems from the 2026 AIME I and II math competitions. Scored by exact answer match. GPT-5.5 100%
  • IMO-AnswerBench In use 400 olympiad problems with short verifiable answers, robustified against memorization. Scored by answer match. GLM-5.2 91%
  • HMMT 2026 In use Models solve the 33 final-answer problems from the February 2026 Harvard-MIT Mathematics Tournament. Scored by exact answer match. GPT-5.5 98.5%
  • PolyMath In use Models solve 9,000 parallel math problems in 18 languages across four difficulty levels, from school math to frontier math. Final answers are checked. Qwen3.7-Max 86.5%
  • MathArena (Apex, IMO, USAMO) In use Evaluates models on newly released math competitions after release. Final-answer or human/LLM-judged proof scores. Claude Opus 4.8 81.3%
  • MathArena Apex Shortlist In use Models solve 47 hard final-answer problems that MathArena selected from 2025 competitions, where the best models scored about 50% at selection. Scored by accuracy over repeated attempts. GPT-5.5 98.4%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 19 months ago, in April 2025, for Qwen3-0.6B. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the benchmark’s paper (28 entries).
  • Pass Models scored 13 models we track, plus 15 other entries on its leaderboard.
  • Warning Newest score The newest came 19 months ago, in April 2025, for Qwen3-0.6B.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Gemini 2.5 Pro (#104).
  • Pass Headroom The best is 74.2%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 13 rows for models we track and 15 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.

How we check