HTTP/2 301 server: GitHub.com content-type: text/html location: https://mathllm.github.io/mathvision/ access-control-allow-origin: * strict-transport-security: max-age=31556952 expires: Tue, 30 Dec 2025 12:04:49 GMT cache-control: max-age=600 x-proxy-cache: MISS x-github-request-id: 0A8C:2D8B9D:A07A20:B43581:6953BD89 accept-ranges: bytes age: 0 date: Tue, 30 Dec 2025 11:54:49 GMT via: 1.1 varnish x-served-by: cache-bom-vanm7210075-BOM x-cache: MISS x-cache-hits: 0 x-timer: S1767095690.653095,VS0,VE204 vary: Accept-Encoding x-fastly-request-id: dc5210e1bf71cab284d7c53a2802d6e871c7e003 content-length: 162 HTTP/2 200 server: GitHub.com content-type: text/html; charset=utf-8 last-modified: Mon, 29 Sep 2025 02:14:37 GMT access-control-allow-origin: * strict-transport-security: max-age=31556952 etag: W/"68d9eb8d-16a52" expires: Tue, 30 Dec 2025 12:04:49 GMT cache-control: max-age=600 content-encoding: gzip x-proxy-cache: MISS x-github-request-id: 6904:2BC55:A087F5:B44384:6953BD89 accept-ranges: bytes age: 0 date: Tue, 30 Dec 2025 11:54:50 GMT via: 1.1 varnish x-served-by: cache-bom-vanm7210075-BOM x-cache: MISS x-cache-hits: 0 x-timer: S1767095690.870371,VS0,VE214 vary: Accept-Encoding x-fastly-request-id: 99a5f7cc3dd6d6ee0d8ea6bdb53aa983c70bf6b3 content-length: 10851

MATH-Vision

Measuring Multimodal Mathematical Reasoning

NeurIPS DB Track, 2024

Paper Code

🏆

Leaderboard

🏆

Open Source Leaderboard

🤗

Dataset

🔮

Visualization

[2024-05-20] (a) Zero-shot accuracies of four prominent Large Multimodal Models (LMMs), random chance, and human performance are evaluated on our proposed MATH-V across 16 subjects. Teal means newly introduced subjects. (b) Examples of easy problems in MATH-V failed by top-performing LMMs on MathVista. The three questions come from tests designed for elementary school students.

[2024-02-21] The accuracies of four prominent Large Multimodal Models (LMMs), random chance, and human performance are evaluated on our proposed MATH-Vision (MATH-V) across 16 subjects and 5 levels of difficulty, with Level 1 being the easiest and Level 5 the most challenging. Human performance is assessed using the testmini subset.

Introduction

Recent advancements in Large Multimodal Models (LMMs) have shown promising results in mathematical reasoning within visual contexts, with models approaching human-level performance on existing benchmarks such as MathVista. However, we observe significant limitations in the diversity of questions and breadth of subjects covered by these benchmarks.

To address this issue, we present the MATH-Vision dataset, a meticulously curated collection of 3,040 high-quality mathematical problems with visual contexts sourced from real math competitions. Spanning 16 distinct mathematical disciplines and graded across 5 levels of difficulty, our dataset provides a comprehensive and diverse set of challenges for evaluating the mathematical reasoning abilities of LMMs.

Through extensive experimentation, we unveil a notable performance gap between current LMMs and human performance on MATH-V, underscoring the imperative for further advancements in LMMs. Moreover, our detailed categorization allows for a thorough error analysis of LMMs, offering valuable insights to guide future research and development.

Official Leaderboard

Accuracy scores on the test set (3,040 examples) of MATH-Vision.

🚨 To submit your results to the leaderboard, please send to this email with your results in this format.
(The format template is on GitHub. If the download fails, please check your network.)

#	Model	Source	Date	ALL	Alg	AnaG	Ari	CombG	Comb	Cnt	DescG	GrphT	Log	Angle	Area	Len	SolG	Stat	Topo	TransG
0	Human	Link	2024-04-05	68.82	55.1	78.6	99.6	98.4	43.5	98.5	91.3	62.2	61.3	33.5	47.2	73.5	87.3	93.1	99.8	69.0
1	Gemini 2.5 Pro 🥇	Link	2025-03-23	73.3	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2	GPT-5 🥈	Link	2025-08-26	72.0	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
3	Seed1.5-VL 🥉	Link	2025-05-12	68.7	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
4	GLM-4.5V (106B-A12B)	Link	2025-08-26	65.6	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
5	Step-3 (321B-A38B)	Link	2025-08-26	64.8	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
6	InternVL3.5 (241B-A28B)	Link	2025-08-26	63.9	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
7	MiMo-VL-RL	Link	2025-06-04	60.4	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
8	OpenAI o1	Link	2025-04-10	60.30	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
9	Step R1-V-Mini	Link	2025-04-05	56.6	58.0	64.3	62.9	43.2	53.6	28.4	33.7	34.4	56.3	66.5	65.8	69.3	53.3	58.6	30.4	46.4
10	InternVL3.5 (30B-A3B)	Link	2025-08-26	55.7	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
11	SenseNova V6 Reasoner	Link	2025-04-10	55.39	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
12	GLM-4.1V (9B)	Link	2025-08-26	54.4	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
13	Kimi k1.6 Preview	Link	2025-03-08	53.29	63.19	54.76	66.43	37.34	51.79	35.82	22.12	34.44	59.66	57.23	57.80	67.04	47.95	55.17	17.39	41.67
14	Open-Vision-Reasoner-7B	Link	2025-07-15	51.8	57.7	41.7	59.3	33.1	50.0	26.9	26.0	40.0	44.5	63.0	60.4	68.4	50.0	51.7	21.7	36.9
15	Skywork-R1V2-38B	Link	2025-04-28	49.7	52.6	47.4	73.7	42.1	52.6	36.8	15.8	57.9	73.7	63.2	73.7	57.9	47.4	47.4	21.1	31.6
16	Doubao-1.5-pro	Link	2025-02-28	48.62	55.07	52.38	63.57	34.74	36.90	43.28	25.00	27.78	37.82	62.43	55.40	59.69	43.85	55.17	26.09	37.50
17	GPT-4.5	Link	2025-04-10	47.30	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
18	VL-Rethinker-72B	Link	2025-03-25	44.93	49.0	48.8	59.3	35.4	33.9	22.4	24.0	32.2	42.9	56.1	50.0	52.8	41.0	65.5	30.4	34.5
19	INFRL-Qwen2.5-VL-72B-Preview	Link	2025-03-25	42.73	49.3	42.9	59.3	31.8	32.7	32.8	22.1	27.8	41.2	54.3	47.6	48.1	38.9	56.9	26.1	33.3
20	Gemini-2 Flash	Link	2025-02-05	41.3	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
21	Kimi k1.5	Link	2025-01-22	38.6	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
22	Qwen2.5-VL-72B	Link	2025-01-26	38.1	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
23	Claude3.5-Sonnet	Link	2024-06-21	37.99	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
24	Kimi-VL-A3B-Thinking	Link	2025-04-11	36.8	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
25	QvQ-72B-Preview	Link	2024-12-25	35.9	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
26	GPT-4o	Link	2024-05-19	30.39	42.0	39.3	49.3	28.9	25.6	22.4	24.0	23.3	29.4	17.3	29.8	30.1	29.1	44.8	34.8	17.9
27	GPT-4 Turbo	Link	2024-05-19	30.26	37.7	33.3	46.4	25.0	28.6	25.3	15.4	27.8	31.9	30.6	29.0	31.9	28.7	37.9	17.4	23.2
28	Claude3-Opus	Link	2024-05-04	27.13	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
29	MathCoder-VL-8B	Link	2025-02-16	26.1	18.6	32.1	26.4	25.0	10.7	13.4	20.2	14.4	21.0	48.6	32.2	32.1	23.0	29.3	8.7	23.2
30	Qwen2-VL-72B	Link	2024-08-29	25.9	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
31	Qwen2.5-VL-7B	Link	2025-01-26	25.1	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
32	TBAC-VLR1-3B-preview	Link	2025-04-21	25.0	22.0	29.8	32.1	19.5	18.5	16.4	22.1	11.1	25.2	39.3	27.6	28.5	22.9	34.5	17.4	22.0
33	CoT GPT4V	Link	2024-02-21	23.98	26.7	26.2	38.6	22.1	24.4	19.4	27.9	23.3	25.2	17.3	21.4	23.4	23.8	25.9	4.4	25.6
34	GPT4V	Link	2024-02-21	22.76	27.3	32.1	35.7	21.1	16.7	13.4	22.1	14.4	16.8	22.0	22.2	20.9	23.8	24.1	21.7	25.6
35	MathCoder-VL-2B	Link	2025-02-16	21.7	15.7	17.9	17.1	19.2	11.3	14.9	26.9	14.4	16.8	38.2	25.4	26.9	15.6	36.2	8.7	25.0
36	Qwen2.5-VL-3B	Link	2025-01-26	21.2	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
37	Gemini-1.5 Pro	Link	2024-05-17	19.24	20.3	35.7	34.3	19.8	15.5	20.9	26.0	26.7	22.7	14.5	14.4	16.5	18.9	10.3	26.1	17.3
38	Ovis1.6-Gemma2-9B	Link	2024-09-19	18.78	13.3	15.5	22.1	17.9	11.3	22.4	23.1	20.0	20.2	20.8	18.0	24.7	15.6	20.7	17.4	20.8
39	Gemini Pro	Link	2024-02-21	17.66	15.1	10.7	20.7	20.1	11.9	7.5	20.2	21.1	16.8	19.1	19.0	20.0	14.3	13.8	17.4	20.8
40	InternVL-Chat-V1-2-Plus	Link	2024-02-22	16.97	11.3	25.0	15.7	16.9	10.1	11.9	16.4	15.6	19.3	22.5	16.4	22.5	14.3	17.2	4.4	20.8
41	Qwen2-VL-7B	Link	2024-08-29	16.3	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
42	Math-LLaVA-13B	Link	2024-06-26	15.69	9.0	20.2	15.7	18.2	10.1	10.5	16.4	14.4	16.0	20.2	18.4	17.6	9.4	24.1	21.7	17.9
43	Qwen-VL-Max	Link	2024-02-21	15.59	10.7	19.1	20.0	16.9	12.5	17.9	16.4	12.2	21.0	13.3	14.2	19.8	11.5	20.7	13.0	17.3
44	InternLM-XComposer2-VL	Link	2024-02-21	14.54	9.3	15.5	12.1	15.3	11.3	10.5	14.4	22.2	19.3	19.7	15.6	15.0	11.9	15.5	26.1	15.5
45	GPT 4-CoT (caption)	Link	2024-02-21	13.10	16.5	20.2	34.3	10.4	17.9	19.4	7.7	11.1	10.1	9.8	9.6	9.1	13.5	13.8	8.7	12.5
46	Qwen2-VL-2B	Link	2024-08-29	12.4	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
47	ShareGPT4V-13B	Link	2024-02-21	11.88	7.5	15.5	16.4	10.7	8.9	9.0	11.5	8.9	7.6	11.6	13.0	17.4	10.3	8.6	8.7	12.5
48	LLaVA-v1.5-13B	Link	2024-02-21	11.12	7.0	14.3	14.3	9.1	6.6	6.0	13.5	5.6	13.5	10.4	12.6	14.7	11.5	13.8	13.0	10.7
49	Qwen-VL-Plus	Link	2024-02-21	10.72	11.3	17.9	14.3	12.7	4.8	10.5	15.4	8.9	14.3	11.6	6.4	10.0	14.3	6.9	8.7	11.31
50	ShareGPT4V-7B	Link	2024-02-21	10.53	5.5	3.6	12.9	10.1	4.8	7.5	11.5	14.4	10.9	16.2	11.8	12.3	9.8	15.5	17.4	11.3
51	SPHINX (V2)	Link	2024-02-21	9.70	6.7	7.1	12.9	7.5	7.7	6.0	9.6	16.7	10.1	11.0	11.8	12.5	8.2	8.6	8.7	6.0
52	LLaVA-v1.5-7B	Link	2024-02-21	8.52	7.0	7.1	10.7	7.1	4.8	10.5	7.7	10.0	9.2	15.6	10.2	9.8	5.3	8.6	4.4	4.8
53	Random Chance	Link	2024-02-21	7.17	1.5	11.9	7.1	9.7	4.8	6.0	22.1	1.1	7.6	0.6	9.4	6.7	8.2	8.6	13.0	7.1

Human*: Average human performance from annotators who have high school diplomas or above.
Subjects: Alg: algebra, AnaG: analytic geometry, Ari: arithmetic, CombG: combinatorial geometry,
Comb: combinatorics, Cnt: counting, DescG: descriptive geometry, GrphT: graph theory, Log: logic,
Angle: metric geometry - angle, Area: metric geometry - area, Len: metric geometry-length,
SolG: solid geometry, Stat: statistics, Topo: topology, TransG: transformation geometry.

Open Source Leaderboard

Accuracy scores on the test set (3,040 examples) of MATH-Vision.

🚨 To submit your results to the leaderboard, please send to this email with your results in this format.
(The format template is on GitHub. If the download fails, please check your network.)

#	Model	Source	Date	ALL	Alg	AnaG	Ari	CombG	Comb	Cnt	DescG	GrphT	Log	Angle	Area	Len	SolG	Stat	Topo	TransG
0	Human	Link	2024-04-05	68.82	55.1	78.6	99.6	98.4	43.5	98.5	91.3	62.2	61.3	33.5	47.2	73.5	87.3	93.1	99.8	69.0
1	Open-Vision-Reasoner-7B 🥇	Link	2025-07-15	51.8	57.7	41.7	59.3	33.1	50.0	26.9	26.0	40.0	44.5	63.0	60.4	68.4	50.0	51.7	21.7	36.9
2	Skywork-R1V2-38B 🥈	Link	2025-04-28	49.7	52.6	47.4	73.7	42.1	52.6	36.8	15.8	57.9	73.7	63.2	73.7	57.9	47.4	47.4	21.1	31.6
3	VL-Rethinker-72B 🥉	Link	2025-03-25	44.93	49.0	48.8	59.3	35.4	33.9	22.4	24.0	32.2	42.9	56.1	50.0	52.8	41.0	65.5	30.4	34.5
4	INFRL-Qwen2.5-VL-72B-Preview	Link	2025-03-25	42.73	49.3	42.9	59.3	31.8	32.7	32.8	22.1	27.8	41.2	54.3	47.6	48.1	38.9	56.9	26.1	33.3
5	Qwen2.5-VL-72B	Link	2025-01-26	38.1	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
6	Kimi-VL-A3B-Thinking	Link	2025-04-11	36.8	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
7	QvQ-72B-Preview	Link	2024-12-25	35.9	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
8	MathCoder-VL-8B	Link	2025-02-16	26.1	18.6	32.1	26.4	25.0	10.7	13.4	20.2	14.4	21.0	48.6	32.2	32.1	23.0	29.3	8.7	23.2
9	Qwen2-VL-72B	Link	2024-08-29	25.9	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
10	Qwen2.5-VL-7B	Link	2025-01-26	25.1	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
11	TBAC-VLR1-3B-preview	Link	2025-04-21	25.0	22.0	29.8	32.1	19.5	18.5	16.4	22.1	11.1	25.2	39.3	27.6	28.5	22.9	34.5	17.4	22.0
12	MathCoder-VL-2B	Link	2025-02-16	21.7	15.7	17.9	17.1	19.2	11.3	14.9	26.9	14.4	16.8	38.2	25.4	26.9	15.6	36.2	8.7	25.0
13	Qwen2.5-VL-3B	Link	2025-01-26	21.2	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
14	Ovis1.6-Gemma2-9B	Link	2024-09-19	18.78	13.3	15.5	22.1	17.9	11.3	22.4	23.1	20.0	20.2	20.8	18.0	24.7	15.6	20.7	17.4	20.8
15	InternVL-Chat-V1-2-Plus	Link	2024-02-22	16.97	11.3	25.0	15.7	16.9	10.1	11.9	16.4	15.6	19.3	22.5	16.4	22.5	14.3	17.2	4.4	20.8
16	Qwen2-VL-7B	Link	2024-08-29	16.3	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
17	Math-LLaVA-13B	Link	2024-06-26	15.69	9.0	20.2	15.7	18.2	10.1	10.5	16.4	14.4	16.0	20.2	18.4	17.6	9.4	24.1	21.7	17.9
18	InternLM-XComposer2-VL	Link	2024-02-21	14.54	9.3	15.5	12.1	15.3	11.3	10.5	14.4	22.2	19.3	19.7	15.6	15.0	11.9	15.5	26.1	15.5
19	Qwen2-VL-2B	Link	2024-08-29	12.4	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
20	ShareGPT4V-13B	Link	2024-02-21	11.88	7.5	15.5	16.4	10.7	8.9	9.0	11.5	8.9	7.6	11.6	13.0	17.4	10.3	8.6	8.7	12.5
21	LLaVA-v1.5-13B	Link	2024-02-21	11.12	7.0	14.3	14.3	9.1	6.6	6.0	13.5	5.6	13.5	10.4	12.6	14.7	11.5	13.8	13.0	10.7
22	ShareGPT4V-7B	Link	2024-02-21	10.53	5.5	3.6	12.9	10.1	4.8	7.5	11.5	14.4	10.9	16.2	11.8	12.3	9.8	15.5	17.4	11.3
23	SPHINX (V2)	Link	2024-02-21	9.70	6.7	7.1	12.9	7.5	7.7	6.0	9.6	16.7	10.1	11.0	11.8	12.5	8.2	8.6	8.7	6.0
24	LLaVA-v1.5-7B	Link	2024-02-21	8.52	7.0	7.1	10.7	7.1	4.8	10.5	7.7	10.0	9.2	15.6	10.2	9.8	5.3	8.6	4.4	4.8
25	Random Chance	Link	2024-02-21	7.17	1.5	11.9	7.1	9.7	4.8	6.0	22.1	1.1	7.6	0.6	9.4	6.7	8.2	8.6	13.0	7.1

Human*: Average human performance from annotators who have high school diplomas or above.
Subjects: Alg: algebra, AnaG: analytic geometry, Ari: arithmetic, CombG: combinatorial geometry,
Comb: combinatorics, Cnt: counting, DescG: descriptive geometry, GrphT: graph theory, Log: logic,
Angle: metric geometry - angle, Area: metric geometry - area, Len: metric geometry-length,
SolG: solid geometry, Stat: statistics, Topo: topology, TransG: transformation geometry.

MATH-V Dataset

Key statistics of MATH-V

Comparison of the level distribution between our
MATH-V and the MATH dataset.

Distribution

levels, subjects and sources distribution of MATH-V.

Visualization

@inproceedings{
wang2025mathcodervl,
title={MathCoder-{VL}: Bridging Vision and Code for Enhanced Multimodal Mathematical Reasoning},
author={Ke Wang and Junting Pan and Linda Wei and Aojun Zhou and Weikang Shi and Zimu Lu and Han Xiao and Yunqiao Yang and Houxing Ren and Mingjie Zhan and Hongsheng Li},
booktitle={The 63rd Annual Meeting of the Association for Computational Linguistics},
year={2025},
url={https://openreview.net/forum?id=nuvtX1imAb}
}
@inproceedings{
wang2024measuring,
title={Measuring Multimodal Mathematical Reasoning with MATH-Vision Dataset},
author={Ke Wang and Junting Pan and Weikang Shi and Zimu Lu and Houxing Ren and Aojun Zhou and Mingjie Zhan and Hongsheng Li},
booktitle={The Thirty-eight Conference on Neural Information Processing Systems Datasets and Benchmarks Track},
year={2024},
url={https://openreview.net/forum?id=QWTCcxMpPA}
}