Open LMM Reasoning Leaderboard
Models solve multimodal reasoning tasks. OpenCompass ranks the models using the component results.
The leaderboard aggregates results from component evaluations.
- Made by
- OpenCompass
- Released
- December 2024
- Kind
- Composite indices
- Level
- University
- Scored
- Settled by events
- Answers
- Short answer
- Tasks
- All public
- Measures
- Capability
- Interaction
- Single turn
- Items from
- Another benchmark
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Vision
- Strongest model tested
- Claude 3.7 Sonnet #119 on the capability index
- Last model added
- April 2025
- Scores refreshed
- November 20th, 2026
Doubao-1.5-pro leads with 61.6 among models released by April 2025 Doubao-1.5-pro leads with 61.6 among models released by April 2025
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. 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:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 61.6
Doubao-1.5-pro
- 56.6
Gemini 2.0 Pro
- 54.8 ChatGPT-4o-latest
- 54.1 GPT-4.1
- 51.0
InternVL3-78B
- 50.6
Gemini 2.0 Flash
- 50.4 Claude 3.7 Sonnet
- 50.3
InternVL3-38B
- 50.2
Qwen2.5-VL-72B
- 50.0
Gemini 1.5 Pro
- 49.1 GLM-4v-Plus-20250111
- 47.9 Claude 3.5 Sonnet
- 47.9
Ovis2-34B
- 46.9
QVQ-72B-Preview
- 46.0
InternVL3-14B
- 46.0 GPT-4.1 mini
- 45.4 Step-1o
- 44.7
Ovis2-16B
- 44.3
InternVL2.5-78B-MPO
- 44.2 GPT-4o (1120, detail-high)
- 43.6 Taichu-VLR-7B
- 43.2
InternVL2.5-38B-MPO
- 42.5 VLAA-Thinker-Qwen2.5VL-7B
- 41.9
InternVL2.5-38B
- 41.7
InternVL2_5-78B
- 41.4 SenseNova
- 41.4
InternVL3-8B
- 40.8
InternVL3-9B
- 40.6
Gemma 3 27B
- 40.1
Qwen2.5-VL-7B
- 38.7 Qwen2-VL-72B
- 38.0 moonshot-v1-8k
- 37.8
Ovis2-8B
- 37.3 InternVL2.5-8B-BoN-8
- 37.0
Ovis1.6-Gemma2-27B
- 37.0 Grok 2 Vision (1212)
- 36.6 MUG-U-7B
- 36.0
SAIL-VL-1.6-8B
- 35.8 Kimi-VL-A3B-Instruct
- 35.4 VLAA-Thinker-Qwen2.5VL-3B
- 34.8 SAIL-VL-1.5-8B
- 34.7 LLaVA-OneVision-72B (SI)
- 34.4
InternVL2-Llama3-76B
- 34.2 GPT-4.1 nano
- 33.9
Gemma 3 12B
- 33.7
NVLM-D-72B
- 33.6 Taichu-VLR-3B
- 33.6 MiniCPM-o-2_6
- 33.4 VLM-R1-3B-Math-0305
- 33.3
Ovis2-4B
- 33.0 Kimi-VL-A3B-Thinking
- 32.9 URSA-8B
- 32.9
InternVL2.5-26B
- 31.8
Qwen2.5-VL-3B
- 31.4 POINTS1.5-Qwen2.5-7B
- 29.8 Yi-Vision
- 29.5 VITA-1.5
- 29.1
InternVL3-2B
- 28.8 Qwen2-VL-7B-Instruct
- 27.9 SAIL-VL-1.5-2B
- 27.6
Ovis2-2B
- 27.1
Gemma 3 4B
- 27.0 LLaVA-CoT
- 26.9
InternVL2-8B
- 26.6
LLaVA-OneVision 7B (SI)
- 26.4
InternVL2-26B
- 25.8
InternLM-XComposer2.5-7B
- 25.2
SAIL-VL-2B
- 24.6 InternVL-Chat-V1-5
- 24.3 Aquila-VL-2B
- 23.0
Ovis2-1B
- 22.5 GLM-4V-9B
- 21.7
Llama 3.2 11B
- 21.7
InternVL2.5-2B-MPO
- 21.3
InternVL3-1B
- 21.2
InternVL2.5-2B
- 20.5
Qwen2-VL-2B-Instruct
- 19.1
Janus-Pro-7B
- 18.5
VILA1.5-13B
- 18.4
SmolVLM-Instruct
- 13.3
LLaVA-v1.5-13B
- 12.1 llava-v1.5-7b
- 1
Doubao-1.5-pro 61.6
- 2
Gemini 2.0 Pro 56.6
- 3 ChatGPT-4o-latest 54.8
- 4 GPT-4.1 54.1
- 5
InternVL3-78B 51.0
- 6
Gemini 2.0 Flash Feb 2025 50.6
- 7 Claude 3.7 Sonnet 50.4
- 8
InternVL3-38B 50.3
- 9
Qwen2.5-VL-72B 50.2
- 10
Gemini 1.5 Pro Sep 2024 50.0
- 11 GLM-4v-Plus-20250111 49.1
- 12 Claude 3.5 Sonnet 47.9
- 13
Ovis2-34B 47.9
- 14
QVQ-72B-Preview 46.9
- 15
InternVL3-14B 46.0
- 16 GPT-4.1 mini 46.0
- 17 Step-1o 45.4
- 18
Ovis2-16B 44.7
- 19
InternVL2.5-78B-MPO 44.3
- 20 GPT-4o (1120, detail-high) 44.2
- 21 Taichu-VLR-7B 43.6
- 22
InternVL2.5-38B-MPO 43.2
- 23 VLAA-Thinker-Qwen2.5VL-7B 42.5
- 24
InternVL2.5-38B 41.9
- 25
InternVL2_5-78B 41.7
- 26 SenseNova 41.4
- 27
InternVL3-8B 41.4
- 28
InternVL3-9B 40.8
- 29
Gemma 3 27B 40.6
- 30
Qwen2.5-VL-7B 40.1
- 31 Qwen2-VL-72B 38.7
- 32 moonshot-v1-8k 38.0
- 33
Ovis2-8B 37.8
- 34 InternVL2.5-8B-BoN-8 37.3
- 35
Ovis1.6-Gemma2-27B 37.0
- 36 Grok 2 Vision (1212) 37.0
- 37 MUG-U-7B 36.6
- 38
SAIL-VL-1.6-8B 36.0
- 39 Kimi-VL-A3B-Instruct 35.8
- 40 VLAA-Thinker-Qwen2.5VL-3B 35.4
- 41 SAIL-VL-1.5-8B 34.8
- 42 LLaVA-OneVision-72B (SI) 34.7
- 43
InternVL2-Llama3-76B 34.4
- 44 GPT-4.1 nano 34.2
- 45
Gemma 3 12B 33.9
- 46
NVLM-D-72B 33.7
- 47 Taichu-VLR-3B 33.6
- 48 MiniCPM-o-2_6 33.6
- 49 VLM-R1-3B-Math-0305 33.4
- 50
Ovis2-4B 33.3
- 51 Kimi-VL-A3B-Thinking 33.0
- 52 URSA-8B 32.9
- 53
InternVL2.5-26B 32.9
- 54
Qwen2.5-VL-3B 31.8
- 55 POINTS1.5-Qwen2.5-7B 31.4
- 56 Yi-Vision 29.8
- 57 VITA-1.5 29.5
- 58
InternVL3-2B 29.1
- 59 Qwen2-VL-7B-Instruct 28.8
- 60 SAIL-VL-1.5-2B 27.9
- 61
Ovis2-2B 27.6
- 62
Gemma 3 4B 27.1
- 63 LLaVA-CoT 27.0
- 64
InternVL2-8B 26.9
- 65
LLaVA-OneVision 7B (SI) 26.6
- 66
InternVL2-26B 26.4
- 67
InternLM-XComposer2.5-7B 25.8
- 68
SAIL-VL-2B 25.2
- 69 InternVL-Chat-V1-5 24.6
- 70 Aquila-VL-2B 24.3
- 71
Ovis2-1B 23.0
- 72 GLM-4V-9B 22.5
- 73
Llama 3.2 11B 21.7
- 74
InternVL2.5-2B-MPO 21.7
- 75
InternVL3-1B 21.3
- 76
InternVL2.5-2B 21.2
- 77
Qwen2-VL-2B-Instruct 20.5
- 78
Janus-Pro-7B 19.1
- 79
VILA1.5-13B 18.5
- 80
SmolVLM-Instruct 18.4
- 81
LLaVA-v1.5-13B 13.3
- 82 llava-v1.5-7b 12.1
Source: Official Open LMM Reasoning Leaderboard leaderboard
- 61.6
Doubao-1.5-pro
- 56.6
Gemini 2.0 Pro
- 54.1 GPT-4.1
- 51.0
InternVL3-78B
- 50.6
Gemini 2.0 Flash
- 50.4 Claude 3.7 Sonnet
- 50.3
InternVL3-38B
- 50.2
Qwen2.5-VL-72B
- 50.0
Gemini 1.5 Pro
- 47.9 Claude 3.5 Sonnet
- 47.9
Ovis2-34B
- 46.9
QVQ-72B-Preview
- 46.0
InternVL3-14B
- 46.0 GPT-4.1 mini
- 44.7
Ovis2-16B
- 44.3
InternVL2.5-78B-MPO
- 43.2
InternVL2.5-38B-MPO
- 41.9
InternVL2.5-38B
- 41.7
InternVL2_5-78B
- 41.4
InternVL3-8B
- 40.8
InternVL3-9B
- 40.6
Gemma 3 27B
- 40.1
Qwen2.5-VL-7B
- 37.8
Ovis2-8B
- 37.0
Ovis1.6-Gemma2-27B
- 37.0 Grok 2 Vision (1212)
- 36.0
SAIL-VL-1.6-8B
- 35.8 Kimi-VL-A3B-Instruct
- 34.4
InternVL2-Llama3-76B
- 34.2 GPT-4.1 nano
- 33.9
Gemma 3 12B
- 33.7
NVLM-D-72B
- 33.6 MiniCPM-o-2_6
- 33.3
Ovis2-4B
- 33.0 Kimi-VL-A3B-Thinking
- 32.9
InternVL2.5-26B
- 31.8
Qwen2.5-VL-3B
- 29.5 VITA-1.5
- 29.1
InternVL3-2B
- 28.8 Qwen2-VL-7B-Instruct
- 27.6
Ovis2-2B
- 27.1
Gemma 3 4B
- 26.9
InternVL2-8B
- 26.6
LLaVA-OneVision 7B (SI)
- 26.4
InternVL2-26B
- 25.8
InternLM-XComposer2.5-7B
- 25.2
SAIL-VL-2B
- 24.6 InternVL-Chat-V1-5
- 24.3 Aquila-VL-2B
- 23.0
Ovis2-1B
- 22.5 GLM-4V-9B
- 21.7
Llama 3.2 11B
- 21.7
InternVL2.5-2B-MPO
- 21.3
InternVL3-1B
- 21.2
InternVL2.5-2B
- 20.5
Qwen2-VL-2B-Instruct
- 19.1
Janus-Pro-7B
- 18.5
VILA1.5-13B
- 18.4
SmolVLM-Instruct
- 13.3
LLaVA-v1.5-13B
- 12.1 llava-v1.5-7b
- 1
Doubao-1.5-pro 61.6
- 2
Gemini 2.0 Pro 56.6
- 3 GPT-4.1 54.1
- 4
InternVL3-78B 51.0
- 5
Gemini 2.0 Flash Feb 2025 50.6
- 6 Claude 3.7 Sonnet 50.4
- 7
InternVL3-38B 50.3
- 8
Qwen2.5-VL-72B 50.2
- 9
Gemini 1.5 Pro Sep 2024 50.0
- 10 Claude 3.5 Sonnet 47.9
- 11
Ovis2-34B 47.9
- 12
QVQ-72B-Preview 46.9
- 13
InternVL3-14B 46.0
- 14 GPT-4.1 mini 46.0
- 15
Ovis2-16B 44.7
- 16
InternVL2.5-78B-MPO 44.3
- 17
InternVL2.5-38B-MPO 43.2
- 18
InternVL2.5-38B 41.9
- 19
InternVL2_5-78B 41.7
- 20
InternVL3-8B 41.4
- 21
InternVL3-9B 40.8
- 22
Gemma 3 27B 40.6
- 23
Qwen2.5-VL-7B 40.1
- 24
Ovis2-8B 37.8
- 25
Ovis1.6-Gemma2-27B 37.0
- 26 Grok 2 Vision (1212) 37.0
- 27
SAIL-VL-1.6-8B 36.0
- 28 Kimi-VL-A3B-Instruct 35.8
- 29
InternVL2-Llama3-76B 34.4
- 30 GPT-4.1 nano 34.2
- 31
Gemma 3 12B 33.9
- 32
NVLM-D-72B 33.7
- 33 MiniCPM-o-2_6 33.6
- 34
Ovis2-4B 33.3
- 35 Kimi-VL-A3B-Thinking 33.0
- 36
InternVL2.5-26B 32.9
- 37
Qwen2.5-VL-3B 31.8
- 38 VITA-1.5 29.5
- 39
InternVL3-2B 29.1
- 40 Qwen2-VL-7B-Instruct 28.8
- 41
Ovis2-2B 27.6
- 42
Gemma 3 4B 27.1
- 43
InternVL2-8B 26.9
- 44
LLaVA-OneVision 7B (SI) 26.6
- 45
InternVL2-26B 26.4
- 46
InternLM-XComposer2.5-7B 25.8
- 47
SAIL-VL-2B 25.2
- 48 InternVL-Chat-V1-5 24.6
- 49 Aquila-VL-2B 24.3
- 50
Ovis2-1B 23.0
- 51 GLM-4V-9B 22.5
- 52
Llama 3.2 11B 21.7
- 53
InternVL2.5-2B-MPO 21.7
- 54
InternVL3-1B 21.3
- 55
InternVL2.5-2B 21.2
- 56
Qwen2-VL-2B-Instruct 20.5
- 57
Janus-Pro-7B 19.1
- 58
VILA1.5-13B 18.5
- 59
SmolVLM-Instruct 18.4
- 60
LLaVA-v1.5-13B 13.3
- 61 llava-v1.5-7b 12.1
Source: Official Open LMM Reasoning Leaderboard leaderboard
| # | Entry | Score | Setting | Date | Link | Source |
|---|---|---|---|---|---|---|
| 1 | | 61.6 | Jan 2025 | team.doubao.com | Official leaderboard | |
| 2 | | 56.6 | Feb 2025 | deepmind.google | Official leaderboard | |
| 3 | ChatGPT-4o-latest | 54.8 | openai.com | Official leaderboard | ||
| 4 | GPT-4.1 | 54.1 | Apr 2025 | openai.com | Official leaderboard | |
| 5 | | 51.0 | Apr 2025 | huggingface.co | Official leaderboard | |
| 6 | | 50.6 | Feb 2025 | deepmind.google | Official leaderboard | |
| 7 | Claude 3.7 Sonnet | 50.4 | Feb 2025 | docs.anthropic.com | Official leaderboard | |
| 8 | | 50.3 | Apr 2025 | huggingface.co | Official leaderboard | |
| 9 | | 50.2 | Jan 2025 | huggingface.co | Official leaderboard | |
| 10 | | 50.0 | Sep 2024 | deepmind.google | Official leaderboard | |
| 11 | GLM-4v-Plus-20250111 | 49.1 | open.bigmodel.cn | Official leaderboard | ||
| 12 | Claude 3.5 Sonnet | 47.9 | Jun 2024 | docs.anthropic.com | Official leaderboard | |
| 13 | | 47.9 | Feb 2025 | huggingface.co | Official leaderboard | |
| 14 | | 46.9 | Dec 2024 | huggingface.co | Official leaderboard | |
| 15 | | 46.0 | Apr 2025 | huggingface.co | Official leaderboard | |
| 16 | GPT-4.1 mini | 46.0 | Apr 2025 | openai.com | Official leaderboard | |
| 17 | Step-1o | 45.4 | platform.stepfun.com | Official leaderboard | ||
| 18 | | 44.7 | Feb 2025 | huggingface.co | Official leaderboard | |
| 19 | | 44.3 | Dec 2024 | huggingface.co | Official leaderboard | |
| 20 | GPT-4o (1120, detail-high) | 44.2 | openai.com | Official leaderboard | ||
| 21 | Taichu-VLR-7B 7B | 43.6 | platform.wair.ac.cn | Official leaderboard | ||
| 22 | | 43.2 | Dec 2024 | huggingface.co | Official leaderboard | |
| 23 | VLAA-Thinker-Qwen2.5VL-7B 8.3B | 42.5 | huggingface.co | Official leaderboard | ||
| 24 | | 41.9 | Dec 2024 | huggingface.co | Official leaderboard | |
| 25 | | 41.7 | Dec 2024 | huggingface.co | Official leaderboard | |
| 26 | SenseNova | 41.4 | platform.sensenova.cn | Official leaderboard | ||
| 27 | | 41.4 | Apr 2025 | huggingface.co | Official leaderboard | |
| 28 | | 40.8 | Apr 2025 | huggingface.co | Official leaderboard | |
| 29 | | 40.6 | Mar 2025 | huggingface.co | Official leaderboard | |
| 30 | | 40.1 | Jan 2025 | huggingface.co | Official leaderboard | |
| 31 | Qwen2-VL-72B 73.4B | 38.7 | huggingface.co | Official leaderboard | ||
| 32 | moonshot-v1-8k | 38.0 | platform.moonshot.cn | Official leaderboard | ||
| 33 | | 37.8 | Feb 2025 | huggingface.co | Official leaderboard | |
| 34 | InternVL2.5-8B-BoN-8 8B | 37.3 | huggingface.co | Official leaderboard | ||
| 35 | | 37.0 | Nov 2024 | huggingface.co | Official leaderboard | |
| 36 | Grok 2 Vision (1212) | 37.0 | Dec 2024 | docs.x.ai | Official leaderboard | |
| 37 | MUG-U-7B 7B | 36.6 | github.com | Official leaderboard | ||
| 38 | | 36.0 | Apr 2025 | huggingface.co | Official leaderboard | |
| 39 | Kimi-VL-A3B-Instruct | 35.8 | Apr 2025 | huggingface.co | Official leaderboard | |
| 40 | VLAA-Thinker-Qwen2.5VL-3B 3.8B | 35.4 | huggingface.co | Official leaderboard | ||
| 41 | SAIL-VL-1.5-8B 8B | 34.8 | huggingface.co | Official leaderboard | ||
| 42 | LLaVA-OneVision-72B (SI) 73B | 34.7 | huggingface.co | Official leaderboard | ||
| 43 | | 34.4 | Jul 2024 | huggingface.co | Official leaderboard | |
| 44 | GPT-4.1 nano | 34.2 | Apr 2025 | openai.com | Official leaderboard | |
| 45 | | 33.9 | Mar 2025 | huggingface.co | Official leaderboard | |
| 46 | | 33.7 | Sep 2024 | huggingface.co | Official leaderboard | |
| 47 | Taichu-VLR-3B 3B | 33.6 | platform.wair.ac.cn | Official leaderboard | ||
| 48 | MiniCPM-o-2_6 | 33.6 | Jan 2025 | huggingface.co | Official leaderboard | |
| 49 | VLM-R1-3B-Math-0305 3.8B | 33.4 | huggingface.co | Official leaderboard | ||
| 50 | | 33.3 | Feb 2025 | huggingface.co | Official leaderboard | |
| 51 | Kimi-VL-A3B-Thinking | 33.0 | Apr 2025 | huggingface.co | Official leaderboard | |
| 52 | URSA-8B 8B | 32.9 | huggingface.co | Official leaderboard | ||
| 53 | | 32.9 | Dec 2024 | huggingface.co | Official leaderboard | |
| 54 | | 31.8 | Jan 2025 | huggingface.co | Official leaderboard | |
| 55 | POINTS1.5-Qwen2.5-7B 8.3B | 31.4 | huggingface.co | Official leaderboard | ||
| 56 | Yi-Vision | 29.8 | platform.lingyiwanwu.com | Official leaderboard | ||
| 57 | VITA-1.5 | 29.5 | Dec 2024 | huggingface.co | Official leaderboard | |
| 58 | | 29.1 | Apr 2025 | huggingface.co | Official leaderboard | |
| 59 | Qwen2-VL-7B-Instruct | 28.8 | Aug 2024 | huggingface.co | Official leaderboard | |
| 60 | SAIL-VL-1.5-2B 2.5B | 27.9 | huggingface.co | Official leaderboard | ||
| 61 | | 27.6 | Feb 2025 | huggingface.co | Official leaderboard | |
| 62 | | 27.1 | Mar 2025 | huggingface.co | Official leaderboard | |
| 63 | LLaVA-CoT 11B | 27.0 | huggingface.co | Official leaderboard | ||
| 64 | | 26.9 | Jul 2024 | huggingface.co | Official leaderboard | |
| 65 | | 26.6 | Aug 2024 | huggingface.co | Official leaderboard | |
| 66 | | 26.4 | Jul 2024 | huggingface.co | Official leaderboard | |
| 67 | | 25.8 | Jul 2024 | huggingface.co | Official leaderboard | |
| 68 | | 25.2 | Dec 2024 | huggingface.co | Official leaderboard | |
| 69 | InternVL-Chat-V1-5 | 24.6 | Apr 2024 | github.com | Official leaderboard | |
| 70 | Aquila-VL-2B | 24.3 | Oct 2024 | huggingface.co | Official leaderboard | |
| 71 | | 23.0 | Feb 2025 | huggingface.co | Official leaderboard | |
| 72 | GLM-4V-9B | 22.5 | Jun 2024 | huggingface.co | Official leaderboard | |
| 73 | | 21.7 | Sep 2024 | huggingface.co | Official leaderboard | |
| 74 | | 21.7 | Dec 2024 | huggingface.co | Official leaderboard | |
| 75 | | 21.3 | Apr 2025 | huggingface.co | Official leaderboard | |
| 76 | | 21.2 | Dec 2024 | huggingface.co | Official leaderboard | |
| 77 | | 20.5 | Aug 2024 | huggingface.co | Official leaderboard | |
| 78 | | 19.1 | Jan 2025 | huggingface.co | Official leaderboard | |
| 79 | | 18.5 | Apr 2024 | huggingface.co | Official leaderboard | |
| 80 | | 18.4 | Nov 2024 | huggingface.co | Official leaderboard | |
| 81 | | 13.3 | Oct 2023 | arxiv.org | Official leaderboard | |
| 82 | llava-v1.5-7b | 12.1 | Oct 2023 | arxiv.org | Official leaderboard |
| # | Model | Score | Setting | Date | Link | Source |
|---|---|---|---|---|---|---|
| 1 | | 61.6 | Jan 2025 | team.doubao.com | Official leaderboard | |
| 2 | | 56.6 | Feb 2025 | deepmind.google | Official leaderboard | |
| 3 | GPT-4.1 | 54.1 | Apr 2025 | openai.com | Official leaderboard | |
| 4 | | 51.0 | Apr 2025 | huggingface.co | Official leaderboard | |
| 5 | | 50.6 | Feb 2025 | deepmind.google | Official leaderboard | |
| 6 | Claude 3.7 Sonnet | 50.4 | Feb 2025 | docs.anthropic.com | Official leaderboard | |
| 7 | | 50.3 | Apr 2025 | huggingface.co | Official leaderboard | |
| 8 | | 50.2 | Jan 2025 | huggingface.co | Official leaderboard | |
| 9 | | 50.0 | Sep 2024 | deepmind.google | Official leaderboard | |
| 10 | Claude 3.5 Sonnet | 47.9 | Jun 2024 | docs.anthropic.com | Official leaderboard | |
| 11 | | 47.9 | Feb 2025 | huggingface.co | Official leaderboard | |
| 12 | | 46.9 | Dec 2024 | huggingface.co | Official leaderboard | |
| 13 | | 46.0 | Apr 2025 | huggingface.co | Official leaderboard | |
| 14 | GPT-4.1 mini | 46.0 | Apr 2025 | openai.com | Official leaderboard | |
| 15 | | 44.7 | Feb 2025 | huggingface.co | Official leaderboard | |
| 16 | | 44.3 | Dec 2024 | huggingface.co | Official leaderboard | |
| 17 | | 43.2 | Dec 2024 | huggingface.co | Official leaderboard | |
| 18 | | 41.9 | Dec 2024 | huggingface.co | Official leaderboard | |
| 19 | | 41.7 | Dec 2024 | huggingface.co | Official leaderboard | |
| 20 | | 41.4 | Apr 2025 | huggingface.co | Official leaderboard | |
| 21 | | 40.8 | Apr 2025 | huggingface.co | Official leaderboard | |
| 22 | | 40.6 | Mar 2025 | huggingface.co | Official leaderboard | |
| 23 | | 40.1 | Jan 2025 | huggingface.co | Official leaderboard | |
| 24 | | 37.8 | Feb 2025 | huggingface.co | Official leaderboard | |
| 25 | | 37.0 | Nov 2024 | huggingface.co | Official leaderboard | |
| 26 | Grok 2 Vision (1212) | 37.0 | Dec 2024 | docs.x.ai | Official leaderboard | |
| 27 | | 36.0 | Apr 2025 | huggingface.co | Official leaderboard | |
| 28 | Kimi-VL-A3B-Instruct | 35.8 | Apr 2025 | huggingface.co | Official leaderboard | |
| 29 | | 34.4 | Jul 2024 | huggingface.co | Official leaderboard | |
| 30 | GPT-4.1 nano | 34.2 | Apr 2025 | openai.com | Official leaderboard | |
| 31 | | 33.9 | Mar 2025 | huggingface.co | Official leaderboard | |
| 32 | | 33.7 | Sep 2024 | huggingface.co | Official leaderboard | |
| 33 | MiniCPM-o-2_6 | 33.6 | Jan 2025 | huggingface.co | Official leaderboard | |
| 34 | | 33.3 | Feb 2025 | huggingface.co | Official leaderboard | |
| 35 | Kimi-VL-A3B-Thinking | 33.0 | Apr 2025 | huggingface.co | Official leaderboard | |
| 36 | | 32.9 | Dec 2024 | huggingface.co | Official leaderboard | |
| 37 | | 31.8 | Jan 2025 | huggingface.co | Official leaderboard | |
| 38 | VITA-1.5 | 29.5 | Dec 2024 | huggingface.co | Official leaderboard | |
| 39 | | 29.1 | Apr 2025 | huggingface.co | Official leaderboard | |
| 40 | Qwen2-VL-7B-Instruct | 28.8 | Aug 2024 | huggingface.co | Official leaderboard | |
| 41 | | 27.6 | Feb 2025 | huggingface.co | Official leaderboard | |
| 42 | | 27.1 | Mar 2025 | huggingface.co | Official leaderboard | |
| 43 | | 26.9 | Jul 2024 | huggingface.co | Official leaderboard | |
| 44 | | 26.6 | Aug 2024 | huggingface.co | Official leaderboard | |
| 45 | | 26.4 | Jul 2024 | huggingface.co | Official leaderboard | |
| 46 | | 25.8 | Jul 2024 | huggingface.co | Official leaderboard | |
| 47 | | 25.2 | Dec 2024 | huggingface.co | Official leaderboard | |
| 48 | InternVL-Chat-V1-5 | 24.6 | Apr 2024 | github.com | Official leaderboard | |
| 49 | Aquila-VL-2B | 24.3 | Oct 2024 | huggingface.co | Official leaderboard | |
| 50 | | 23.0 | Feb 2025 | huggingface.co | Official leaderboard | |
| 51 | GLM-4V-9B | 22.5 | Jun 2024 | huggingface.co | Official leaderboard | |
| 52 | | 21.7 | Sep 2024 | huggingface.co | Official leaderboard | |
| 53 | | 21.7 | Dec 2024 | huggingface.co | Official leaderboard | |
| 54 | | 21.3 | Apr 2025 | huggingface.co | Official leaderboard | |
| 55 | | 21.2 | Dec 2024 | huggingface.co | Official leaderboard | |
| 56 | | 20.5 | Aug 2024 | huggingface.co | Official leaderboard | |
| 57 | | 19.1 | Jan 2025 | huggingface.co | Official leaderboard | |
| 58 | | 18.5 | Apr 2024 | huggingface.co | Official leaderboard | |
| 59 | | 18.4 | Nov 2024 | huggingface.co | Official leaderboard | |
| 60 | | 13.3 | Oct 2023 | arxiv.org | Official leaderboard | |
| 61 | llava-v1.5-7b | 12.1 | Oct 2023 | arxiv.org | Official leaderboard |
Rankings by part
We read 6 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as MathVista, MathVision and MathVerse. Doubao-1.5-pro leads 2 of them, and 3 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
MathVista
61 models we track have a score here.
- 1
InternVL3-78B ↑3 79.0
- 2
Doubao-1.5-pro ↓1 78.6
- 3
InternVL2.5-78B-MPO ↑13 76.6
- 4
InternVL3-38B ↑3 76.3
- 5
Ovis2-34B ↑6 76.1
- 6
InternVL3-14B ↑7 74.4
- 7
Qwen2.5-VL-72B ↑1 74.2
- 8
SAIL-VL-1.6-8B ↑19 74.2
- 9
Ovis2-16B ↑6 73.7
- 10
InternVL2.5-38B-MPO ↑7 73.6
MathVision
61 models we track have a score here.
- 1
Doubao-1.5-pro 51.5
- 2
Gemini 2.0 Pro 48.1
- 3 GPT-4.1 mini ↑11 45.3
- 4 GPT-4.1 ↓1 45.1
- 5
Gemini 2.0 Flash Feb 2025 43.6
- 6 Claude 3.7 Sonnet 41.9
- 7
Gemini 1.5 Pro Sep 2024 ↑2 41.0
- 8
Qwen2.5-VL-72B 39.3
- 9
InternVL3-78B ↓5 38.8
- 10
InternVL2.5-38B-MPO ↑7 36.8
MathVerse
61 models we track have a score here.
- 1
Gemini 2.0 Pro ↑1 67.3
- 2
Doubao-1.5-pro ↓1 64.7
- 3
Gemini 1.5 Pro Sep 2024 ↑6 54.8
- 4
InternVL3-78B 51.0
- 5
Ovis2-34B ↑6 50.1
- 6 GPT-4.1 ↓3 48.9
- 7
QVQ-72B-Preview ↑5 48.2
- 8
InternVL3-38B ↓1 48.2
- 9
Gemini 2.0 Flash Feb 2025 ↓4 47.8
- 10
Qwen2.5-VL-72B ↓2 47.3
DynaMath
61 models we track have a score here.
- 1 GPT-4.1 mini ↑13 47.7
- 2
Doubao-1.5-pro ↓1 44.9
- 3
Gemini 2.0 Pro ↓1 43.3
- 4 GPT-4.1 ↓1 43.3
- 5
Gemini 2.0 Flash Feb 2025 42.1
- 6 Claude 3.7 Sonnet 39.7
- 7
Qwen2.5-VL-72B ↑1 35.9
- 8 Claude 3.5 Sonnet ↑2 35.7
- 9
InternVL3-78B ↓5 35.1
- 10
InternVL3-38B ↓3 34.7
WeMath
61 models we track have a score here.
- 1
Doubao-1.5-pro 65.7
- 2
Gemini 2.0 Pro 56.5
- 3 GPT-4.1 55.5
- 4
Ovis2-34B ↑7 51.9
- 5
Gemini 1.5 Pro Sep 2024 ↑4 50.5
- 6 Claude 3.7 Sonnet 49.3
- 7
InternVL3-38B 49.3
- 8
Qwen2.5-VL-72B 49.1
- 9
Gemini 2.0 Flash Feb 2025 ↓4 47.4
- 10
InternVL3-78B ↓6 46.1
LogicVista
61 models we track have a score here.
- 1 GPT-4.1 mini ↑13 64.9
- 2
Doubao-1.5-pro ↓1 64.2
- 3 GPT-4.1 61.1
- 4 Claude 3.5 Sonnet ↑6 60.4
- 5
QVQ-72B-Preview ↑7 58.2
- 6 Claude 3.7 Sonnet 58.2
- 7
InternVL3-38B 57.7
- 8
InternVL3-78B ↓4 55.9
- 9
Qwen2.5-VL-72B ↓1 55.7
- 10
Gemini 1.5 Pro Sep 2024 ↓1 54.4
2 models have held number one on the official leaderboard since its release in December 2024
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official Open LMM Reasoning Leaderboard 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 Open LMM Reasoning Leaderboard 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 Other
- 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. Claude Fable 5.2 64.6
- 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.
- Vals Science Index In use Agents are scored on the equal average of BioMysteryBench, MysteryMechanism and Terminal-Bench-Science. GPT-6.2 Astra 68.4%
- Decision Index Public v0.3 In use Decision models return probabilities over labeled options in public tasks covering knowledge, language, retrieval, tools, and arts. A chance-corrected weighted aggregate combines the task scores. Granite Switch 4.1 30B 51.3
- SAGE v1.1 (Vals) In use Models grade handwritten undergraduate math and computer science work against rubrics. Scored by class-balanced accuracy of each rubric check against expert grading. Claude Opus 5.5 91.3%
- MTEB Fading Massive Text Embedding Benchmark: retrieval, clustering, classification across languages and modalities.
Qwen3-Embedding-8B 72.2%
Data audit
8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 19 months ago, in April 2025, for SAIL-VL-1.6-8B. 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 (82 entries).
- Pass Models scored 61 models we track, plus 21 other entries on its leaderboard.
- Warning Newest score The newest came 19 months ago, in April 2025, for SAIL-VL-1.6-8B.
- Note Frontier models None of the top 10 has a score. The strongest tested is Claude 3.7 Sonnet (#119).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on huggingface.co: 61 rows for models we track and 21 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.