Open FlagEval VLM Leaderboard
Models answer questions about images. The leaderboard compares their scores across visual understanding tasks.
The publisher lists 52 model directories in the public leaderboard results.
- Released
- February 2025
- Kind
- General visual questions
- Level
- University
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- All public
- Measures
- Capability
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Vision
- Strongest model tested
-
Gemini 2.5 Pro Mar 2025 #104 on the capability index
- Last model added
- April 2025
- Scores refreshed
- November 21st, 2026
Gemini 2.5 Pro leads with 70.7 among models released by April 2025 Gemini 2.5 Pro leads with 70.7 among models released by April 2025
Best score per entry from the benchmark's own leaderboard, read on November 21st, 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
- 70.7
Gemini 2.5 Pro
- 61.9
Gemini 2.0 Pro
- 60.5
Gemini 2.0 Flash
- 60.1 Llama-4-maverick-instruct-basic
- 57.8
InternVL3-78B
- 57.6
Qwen2.5-VL-72B
- 57.5 Claude 3.7 Sonnet
- 57.0
Gemini 1.5 Pro
- 56.7 Claude 3.5 Sonnet
- 56.6
Qwen2.5-VL-32B
- 56.3 Doubao-Pro-Vision-32k-241028
- 55.2 Qwen-VL-Max
- 55.1 GPT-4o
- 54.0 Qwen2-VL-72B-Instruct
- 52.7 GPT-4o
- 50.5
InternVL3-8B
- 49.9 Mistral-Small-3.1-24B-Instruct-2503
- 49.8 Yi-Vision
- 48.8 Step-1V-32k
- 48.8 Pixtral-Large-Instruct-2411
- 48.4
Qwen2.5-VL-7B
- 48.2
InternVL2.5-26B
- 46.6
InternVL2-Llama3-76B
- 46.6 GLM-4V-Plus
- 46.3 LLaVA-OneVision 72B
- 46.1
Gemma 3 27B
- 45.7
InternVL2.5-8B
- 44.9 Qwen2-VL-7B-Instruct
- 44.6
Gemini 1.5 Flash
- 44.5 GPT-4o mini
- 44.0
NVLM-D-72B
- 43.2
Molmo-72B
- 41.3 MiniCPM-V-2_6
- 41.2
Llama 3.2 90B
- 41.0
InternVL2-8B
- 40.9 Claude 3 Opus
- 40.9 Aria
- 38.8
Phi-4-multimodal
- 37.9 Pixtral-12B-2409
- 36.9
Qwen2-VL-2B-Instruct
- 35.7
InternVL2.5-2B
- 35.3
Molmo-7B-D
- 34.6
LLaVA-OneVision 7B
- 32.4 Phi-3.5-vision-instruct
- 31.4
InternVL2-2B
- 30.9
Idefics3-8B-Llama3
- 29.1 XGen-MM-Instruct-Interleave-v1.5
- 26.2
LLaVA-OneVision 0.5B
- 24.9 Mono-InternVL-2B
- 24.0
Llama 3.2 11B
- 21.7
Janus-1.3B
- 1
Gemini 2.5 Pro Mar 2025 70.7
- 2
Gemini 2.0 Pro 61.9
- 3
Gemini 2.0 Flash Feb 2025 60.5
- 4 Llama-4-maverick-instruct-basic 60.1
- 5
InternVL3-78B 57.8
- 6
Qwen2.5-VL-72B 57.6
- 7 Claude 3.7 Sonnet 57.5
- 8
Gemini 1.5 Pro Sep 2024 57.0
- 9 Claude 3.5 Sonnet 56.7
- 10
Qwen2.5-VL-32B 56.6
- 11 Doubao-Pro-Vision-32k-241028 56.3
- 12 Qwen-VL-Max 55.2
- 13 GPT-4o Nov 2024 55.1
- 14 Qwen2-VL-72B-Instruct 54.0
- 15 GPT-4o Aug 2024 52.7
- 16
InternVL3-8B 50.5
- 17 Mistral-Small-3.1-24B-Instruct-2503 49.9
- 18 Yi-Vision 49.8
- 19 Step-1V-32k 48.8
- 20 Pixtral-Large-Instruct-2411 48.8
- 21
Qwen2.5-VL-7B 48.4
- 22
InternVL2.5-26B 48.2
- 23
InternVL2-Llama3-76B 46.6
- 24 GLM-4V-Plus 46.6
- 25 LLaVA-OneVision 72B 46.3
- 26
Gemma 3 27B 46.1
- 27
InternVL2.5-8B 45.7
- 28 Qwen2-VL-7B-Instruct 44.9
- 29
Gemini 1.5 Flash Sep 2024 44.6
- 30 GPT-4o mini 44.5
- 31
NVLM-D-72B 44.0
- 32
Molmo-72B 43.2
- 33 MiniCPM-V-2_6 41.3
- 34
Llama 3.2 90B 41.2
- 35
InternVL2-8B 41.0
- 36 Claude 3 Opus 40.9
- 37 Aria 40.9
- 38
Phi-4-multimodal 38.8
- 39 Pixtral-12B-2409 37.9
- 40
Qwen2-VL-2B-Instruct 36.9
- 41
InternVL2.5-2B 35.7
- 42
Molmo-7B-D 35.3
- 43
LLaVA-OneVision 7B 34.6
- 44 Phi-3.5-vision-instruct 32.4
- 45
InternVL2-2B 31.4
- 46
Idefics3-8B-Llama3 30.9
- 47 XGen-MM-Instruct-Interleave-v1.5 29.1
- 48
LLaVA-OneVision 0.5B 26.2
- 49 Mono-InternVL-2B 24.9
- 50
Llama 3.2 11B 24.0
- 51
Janus-1.3B 21.7
Source: Official Open FlagEval VLM Leaderboard leaderboard
- 70.7
Gemini 2.5 Pro
- 61.9
Gemini 2.0 Pro
- 60.5
Gemini 2.0 Flash
- 57.8
InternVL3-78B
- 57.6
Qwen2.5-VL-72B
- 57.5 Claude 3.7 Sonnet
- 57.0
Gemini 1.5 Pro
- 56.7 Claude 3.5 Sonnet
- 56.6
Qwen2.5-VL-32B
- 55.2 Qwen-VL-Max
- 55.1 GPT-4o
- 54.0 Qwen2-VL-72B-Instruct
- 52.7 GPT-4o
- 50.5
InternVL3-8B
- 48.4
Qwen2.5-VL-7B
- 48.2
InternVL2.5-26B
- 46.6
InternVL2-Llama3-76B
- 46.3 LLaVA-OneVision 72B
- 46.1
Gemma 3 27B
- 45.7
InternVL2.5-8B
- 44.9 Qwen2-VL-7B-Instruct
- 44.6
Gemini 1.5 Flash
- 44.5 GPT-4o mini
- 44.0
NVLM-D-72B
- 43.2
Molmo-72B
- 41.3 MiniCPM-V-2_6
- 41.2
Llama 3.2 90B
- 41.0
InternVL2-8B
- 40.9 Claude 3 Opus
- 40.9 Aria
- 38.8
Phi-4-multimodal
- 36.9
Qwen2-VL-2B-Instruct
- 35.7
InternVL2.5-2B
- 35.3
Molmo-7B-D
- 34.6
LLaVA-OneVision 7B
- 32.4 Phi-3.5-vision-instruct
- 31.4
InternVL2-2B
- 30.9
Idefics3-8B-Llama3
- 26.2
LLaVA-OneVision 0.5B
- 24.0
Llama 3.2 11B
- 21.7
Janus-1.3B
- 1
Gemini 2.5 Pro Mar 2025 70.7
- 2
Gemini 2.0 Pro 61.9
- 3
Gemini 2.0 Flash Feb 2025 60.5
- 4
InternVL3-78B 57.8
- 5
Qwen2.5-VL-72B 57.6
- 6 Claude 3.7 Sonnet 57.5
- 7
Gemini 1.5 Pro Sep 2024 57.0
- 8 Claude 3.5 Sonnet 56.7
- 9
Qwen2.5-VL-32B 56.6
- 10 Qwen-VL-Max 55.2
- 11 GPT-4o Nov 2024 55.1
- 12 Qwen2-VL-72B-Instruct 54.0
- 13 GPT-4o Aug 2024 52.7
- 14
InternVL3-8B 50.5
- 15
Qwen2.5-VL-7B 48.4
- 16
InternVL2.5-26B 48.2
- 17
InternVL2-Llama3-76B 46.6
- 18 LLaVA-OneVision 72B 46.3
- 19
Gemma 3 27B 46.1
- 20
InternVL2.5-8B 45.7
- 21 Qwen2-VL-7B-Instruct 44.9
- 22
Gemini 1.5 Flash Sep 2024 44.6
- 23 GPT-4o mini 44.5
- 24
NVLM-D-72B 44.0
- 25
Molmo-72B 43.2
- 26 MiniCPM-V-2_6 41.3
- 27
Llama 3.2 90B 41.2
- 28
InternVL2-8B 41.0
- 29 Claude 3 Opus 40.9
- 30 Aria 40.9
- 31
Phi-4-multimodal 38.8
- 32
Qwen2-VL-2B-Instruct 36.9
- 33
InternVL2.5-2B 35.7
- 34
Molmo-7B-D 35.3
- 35
LLaVA-OneVision 7B 34.6
- 36 Phi-3.5-vision-instruct 32.4
- 37
InternVL2-2B 31.4
- 38
Idefics3-8B-Llama3 30.9
- 39
LLaVA-OneVision 0.5B 26.2
- 40
Llama 3.2 11B 24.0
- 41
Janus-1.3B 21.7
Source: Official Open FlagEval VLM Leaderboard leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 70.7 | Mar 2025 | Official leaderboard | |
| 2 | | 61.9 | Feb 2025 | Official leaderboard | |
| 3 | | 60.5 | Feb 2025 | Official leaderboard | |
| 4 | Llama-4-maverick-instruct-basic | 60.1 | Official leaderboard | ||
| 5 | | 57.8 | Apr 2025 | Official leaderboard | |
| 6 | | 57.6 | Jan 2025 | Official leaderboard | |
| 7 | Claude 3.7 Sonnet | 57.5 | Feb 2025 | Official leaderboard | |
| 8 | | 57.0 | Sep 2024 | Official leaderboard | |
| 9 | Claude 3.5 Sonnet | 56.7 | Jun 2024 | Official leaderboard | |
| 10 | | 56.6 | Mar 2025 | Official leaderboard | |
| 11 | Doubao-Pro-Vision-32k-241028 | 56.3 | Official leaderboard | ||
| 12 | Qwen-VL-Max | 55.2 | Jan 2024 | Official leaderboard | |
| 13 | GPT-4o Nov 2024 | 55.1 | Nov 2024 | Official leaderboard | |
| 14 | Qwen2-VL-72B-Instruct | 54.0 | Aug 2024 | Official leaderboard | |
| 15 | GPT-4o Aug 2024 | 52.7 | Aug 2024 | Official leaderboard | |
| 16 | | 50.5 | Apr 2025 | Official leaderboard | |
| 17 | Mistral-Small-3.1-24B-Instruct-2503 | 49.9 | Official leaderboard | ||
| 18 | Yi-Vision | 49.8 | Official leaderboard | ||
| 19 | Step-1V-32k | 48.8 | Official leaderboard | ||
| 20 | Pixtral-Large-Instruct-2411 | 48.8 | Official leaderboard | ||
| 21 | | 48.4 | Jan 2025 | Official leaderboard | |
| 22 | | 48.2 | Dec 2024 | Official leaderboard | |
| 23 | | 46.6 | Jul 2024 | Official leaderboard | |
| 24 | GLM-4V-Plus | 46.6 | Official leaderboard | ||
| 25 | LLaVA-OneVision 72B | 46.3 | Aug 2024 | Official leaderboard | |
| 26 | | 46.1 | Mar 2025 | Official leaderboard | |
| 27 | | 45.7 | Dec 2024 | Official leaderboard | |
| 28 | Qwen2-VL-7B-Instruct | 44.9 | Aug 2024 | Official leaderboard | |
| 29 | | 44.6 | Aug 2024 | Official leaderboard | |
| 30 | GPT-4o mini | 44.5 | Jul 2024 | Official leaderboard | |
| 31 | | 44.0 | Sep 2024 | Official leaderboard | |
| 32 | | 43.2 | Sep 2024 | Official leaderboard | |
| 33 | MiniCPM-V-2_6 | 41.3 | Aug 2024 | Official leaderboard | |
| 34 | | 41.2 | Sep 2024 | Official leaderboard | |
| 35 | | 41.0 | Jul 2024 | Official leaderboard | |
| 36 | Claude 3 Opus | 40.9 | Feb 2024 | Official leaderboard | |
| 37 | Aria | 40.9 | Sep 2024 | Official leaderboard | |
| 38 | | 38.8 | Feb 2025 | Official leaderboard | |
| 39 | Pixtral-12B-2409 | 37.9 | Official leaderboard | ||
| 40 | | 36.9 | Aug 2024 | Official leaderboard | |
| 41 | | 35.7 | Dec 2024 | Official leaderboard | |
| 42 | | 35.3 | Sep 2024 | Official leaderboard | |
| 43 | | 34.6 | Aug 2024 | Official leaderboard | |
| 44 | Phi-3.5-vision-instruct | 32.4 | Aug 2024 | Official leaderboard | |
| 45 | | 31.4 | Jul 2024 | Official leaderboard | |
| 46 | | 30.9 | Aug 2024 | Official leaderboard | |
| 47 | XGen-MM-Instruct-Interleave-v1.5 | 29.1 | Official leaderboard | ||
| 48 | | 26.2 | Aug 2024 | Official leaderboard | |
| 49 | Mono-InternVL-2B | 24.9 | Official leaderboard | ||
| 50 | | 24.0 | Sep 2024 | Official leaderboard | |
| 51 | | 21.7 | Oct 2024 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 70.7 | Mar 2025 | Official leaderboard | |
| 2 | | 61.9 | Feb 2025 | Official leaderboard | |
| 3 | | 60.5 | Feb 2025 | Official leaderboard | |
| 4 | | 57.8 | Apr 2025 | Official leaderboard | |
| 5 | | 57.6 | Jan 2025 | Official leaderboard | |
| 6 | Claude 3.7 Sonnet | 57.5 | Feb 2025 | Official leaderboard | |
| 7 | | 57.0 | Sep 2024 | Official leaderboard | |
| 8 | Claude 3.5 Sonnet | 56.7 | Jun 2024 | Official leaderboard | |
| 9 | | 56.6 | Mar 2025 | Official leaderboard | |
| 10 | Qwen-VL-Max | 55.2 | Jan 2024 | Official leaderboard | |
| 11 | GPT-4o Nov 2024 | 55.1 | Nov 2024 | Official leaderboard | |
| 12 | Qwen2-VL-72B-Instruct | 54.0 | Aug 2024 | Official leaderboard | |
| 13 | GPT-4o Aug 2024 | 52.7 | Aug 2024 | Official leaderboard | |
| 14 | | 50.5 | Apr 2025 | Official leaderboard | |
| 15 | | 48.4 | Jan 2025 | Official leaderboard | |
| 16 | | 48.2 | Dec 2024 | Official leaderboard | |
| 17 | | 46.6 | Jul 2024 | Official leaderboard | |
| 18 | LLaVA-OneVision 72B | 46.3 | Aug 2024 | Official leaderboard | |
| 19 | | 46.1 | Mar 2025 | Official leaderboard | |
| 20 | | 45.7 | Dec 2024 | Official leaderboard | |
| 21 | Qwen2-VL-7B-Instruct | 44.9 | Aug 2024 | Official leaderboard | |
| 22 | | 44.6 | Aug 2024 | Official leaderboard | |
| 23 | GPT-4o mini | 44.5 | Jul 2024 | Official leaderboard | |
| 24 | | 44.0 | Sep 2024 | Official leaderboard | |
| 25 | | 43.2 | Sep 2024 | Official leaderboard | |
| 26 | MiniCPM-V-2_6 | 41.3 | Aug 2024 | Official leaderboard | |
| 27 | | 41.2 | Sep 2024 | Official leaderboard | |
| 28 | | 41.0 | Jul 2024 | Official leaderboard | |
| 29 | Claude 3 Opus | 40.9 | Feb 2024 | Official leaderboard | |
| 30 | Aria | 40.9 | Sep 2024 | Official leaderboard | |
| 31 | | 38.8 | Feb 2025 | Official leaderboard | |
| 32 | | 36.9 | Aug 2024 | Official leaderboard | |
| 33 | | 35.7 | Dec 2024 | Official leaderboard | |
| 34 | | 35.3 | Sep 2024 | Official leaderboard | |
| 35 | | 34.6 | Aug 2024 | Official leaderboard | |
| 36 | Phi-3.5-vision-instruct | 32.4 | Aug 2024 | Official leaderboard | |
| 37 | | 31.4 | Jul 2024 | Official leaderboard | |
| 38 | | 30.9 | Aug 2024 | Official leaderboard | |
| 39 | | 26.2 | Aug 2024 | Official leaderboard | |
| 40 | | 24.0 | Sep 2024 | Official leaderboard | |
| 41 | | 21.7 | Oct 2024 | Official leaderboard |
Rankings by part
We read 12 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as CMMMU, MMMU and MMMU_Pro_standard. Gemini 2.5 Pro leads every one of them. Arrows show how far a model moved from its overall rank among the models we track.
CMMMU
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 71.2
- 2
Gemini 2.0 Pro 61.1
- 3
Qwen2.5-VL-72B ↑2 56.6
- 4
Gemini 2.0 Flash Feb 2025 ↓1 54.7
- 5
InternVL3-78B ↓1 52.9
- 6 Qwen2-VL-72B-Instruct ↑6 52.8
- 7 Claude 3.7 Sonnet ↓1 51.1
- 8 Claude 3.5 Sonnet 50.8
- 9 GPT-4o Aug 2024 ↑4 50.8
- 10
NVLM-D-72B ↑14 50.2
MMMU
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 69.6
- 2
InternVL3-78B ↑2 66.4
- 3
Gemini 2.0 Flash Feb 2025 65.9
- 4
Gemini 2.0 Pro ↓2 62.9
- 5
Qwen2.5-VL-72B 62.6
- 6
Gemini 1.5 Pro Sep 2024 ↑1 61.8
- 7
Qwen2.5-VL-32B ↑2 61.0
- 8 Qwen2-VL-72B-Instruct ↑4 60.9
- 9 GPT-4o Nov 2024 ↑2 60.4
- 10
NVLM-D-72B ↑14 58.2
MMMU_Pro_standard
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 56.8
- 2
Gemini 2.0 Flash Feb 2025 ↑1 51.2
- 3
InternVL3-78B ↑1 49.6
- 4
Gemini 1.5 Pro Sep 2024 ↑3 47.8
- 5
Qwen2.5-VL-72B 46.9
- 6
Gemini 2.0 Pro ↓4 45.9
- 7
Qwen2.5-VL-32B ↑2 44.3
- 8 GPT-4o Nov 2024 ↑3 43.2
- 9 Claude 3.5 Sonnet ↓1 42.1
- 10 Qwen2-VL-72B-Instruct ↑2 41.3
MMMU_Pro_vision
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 65.8
- 2 Claude 3.5 Sonnet ↑6 47.6
- 3
Gemini 2.0 Flash Feb 2025 47.6
- 4
Qwen2.5-VL-72B ↑1 46.3
- 5 Claude 3.7 Sonnet ↑1 46.1
- 6 GPT-4o Nov 2024 ↑5 45.0
- 7
Gemini 1.5 Pro Sep 2024 43.2
- 8 Qwen-VL-Max ↑2 41.2
- 9
Gemini 2.0 Pro ↓7 41.0
- 10 GPT-4o Aug 2024 ↑3 39.4
MathVision
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 64.5
- 2
Gemini 2.0 Pro 52.9
- 3
Gemini 2.0 Flash Feb 2025 47.5
- 4
Gemini 1.5 Pro Sep 2024 ↑3 44.5
- 5 Claude 3.7 Sonnet ↑1 44.2
- 6
InternVL3-78B ↓2 40.0
- 7 Claude 3.5 Sonnet ↑1 38.3
- 8
Gemma 3 27B ↑11 33.2
- 9 Qwen-VL-Max ↑1 33.1
- 10
Qwen2.5-VL-32B ↓1 31.2
CII-Bench
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 73.5
- 2 Qwen2-VL-72B-Instruct ↑10 67.8
- 3
Gemini 2.0 Pro ↓1 67.7
- 4
InternVL3-78B 67.3
- 5
Qwen2.5-VL-72B 63.3
- 6 Claude 3.5 Sonnet ↑2 63.1
- 7
Qwen2.5-VL-32B ↑2 63.0
- 8
Gemini 2.0 Flash Feb 2025 ↓5 62.6
- 9
InternVL2.5-26B ↑7 61.8
- 10 GPT-4o Nov 2024 ↑1 61.0
Blink
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 71.8
- 2 GPT-4o Nov 2024 ↑9 65.7
- 3
Gemini 2.0 Pro ↓1 64.7
- 4 Claude 3.7 Sonnet ↑2 64.4
- 5 GPT-4o Aug 2024 ↑8 64.2
- 6
Gemini 1.5 Pro Sep 2024 ↑1 64.1
- 7
Gemini 2.0 Flash Feb 2025 ↓4 62.0
- 8
Phi-4-multimodal ↑23 61.4
- 9
InternVL3-78B ↓5 61.1
- 10 Claude 3.5 Sonnet ↓2 60.3
CharXiv
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 77.6
- 2 Claude 3.5 Sonnet ↑6 71.9
- 3 Claude 3.7 Sonnet ↑3 71.2
- 4
Gemini 2.0 Pro ↓2 67.1
- 5
Gemini 2.0 Flash Feb 2025 ↓2 66.1
- 6 GPT-4o Nov 2024 ↑5 65.5
- 7
Qwen2.5-VL-72B ↓2 65.4
- 8 Qwen-VL-Max ↑2 64.3
- 9
InternVL3-78B ↓5 62.6
- 10
Gemini 1.5 Pro Sep 2024 ↓3 62.4
MathVerse
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 60.3
- 2
Gemini 1.5 Pro Sep 2024 ↑5 57.0
- 3 Claude 3.7 Sonnet ↑3 53.8
- 4
Qwen2.5-VL-32B ↑5 53.3
- 5
Gemini 2.0 Flash Feb 2025 ↓2 53.2
- 6 Claude 3.5 Sonnet ↑2 52.2
- 7 Qwen-VL-Max ↑3 52.0
- 8
Gemini 2.0 Pro ↓6 50.6
- 9
InternVL3-78B ↓5 47.6
- 10
Qwen2.5-VL-7B ↑5 45.4
MmvetV2
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 77.7
- 2
Gemini 2.0 Flash Feb 2025 ↑1 72.4
- 3
Gemini 2.0 Pro ↓1 72.2
- 4 Qwen-VL-Max ↑6 71.0
- 5
Qwen2.5-VL-32B ↑4 70.7
- 6 GPT-4o Nov 2024 ↑5 69.6
- 7
Qwen2.5-VL-72B ↓2 69.2
- 8
InternVL3-78B ↓4 69.1
- 9 Claude 3.7 Sonnet ↓3 68.9
- 10
Gemma 3 27B ↑9 66.2
Ocrlite
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 83.5
- 2
Gemini 2.0 Pro 82.8
- 3
Gemini 2.0 Flash Feb 2025 81.9
- 4 GPT-4o Nov 2024 ↑7 79.5
- 5
Qwen2.5-VL-72B 79.4
- 6 GPT-4o Aug 2024 ↑7 79.2
- 7
Gemini 1.5 Pro Sep 2024 77.7
- 8
Qwen2.5-VL-32B ↑1 77.6
- 9 Qwen-VL-Max ↑1 77.1
- 10
Gemini 1.5 Flash Sep 2024 ↑12 77.1
OcrliteZh
41 models we track have a score here.
- 1
Gemini 2.5 Pro Mar 2025 76.1
- 2
Qwen2.5-VL-72B ↑3 74.8
- 3 Qwen-VL-Max ↑7 73.8
- 4
Gemini 2.0 Pro ↓2 73.7
- 5
Qwen2.5-VL-32B ↑4 71.8
- 6
Qwen2.5-VL-7B ↑9 68.8
- 7 Qwen2-VL-72B-Instruct ↑5 66.7
- 8
InternVL3-78B ↓4 65.4
- 9 Qwen2-VL-7B-Instruct ↑12 61.5
- 10 GPT-4o Aug 2024 ↑3 61.1
3 models have held number one on the official leaderboard since its release in February 2025
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official Open FlagEval VLM 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 FlagEval VLM 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 Vision
- MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy.
Qwen 3.6 Plus 86%
- MMMU-Pro In use Harder MMMU: college-level multimodal questions with more options and vision-only inputs to remove text shortcuts. GPT-5.6 Sol 84.6%
- CharXiv In use Chart understanding on real arXiv figures. Reasoning questions require synthesizing information from complex charts. Claude Fable 5 93.5%
- MathVista In use Math reasoning in visual contexts (charts, geometry, puzzles) from 31 datasets. Answer accuracy.
Seed2.1 Pro 90.7%
- MATH-Vision (MathVision) In use 3,040 competition math problems with visual context. Answer accuracy.
Qwen 3.8 Max 97.7%
- MMStar In use Models answer 1,500 human-curated image questions that cannot be solved without the image, across 6 core capabilities. Scored by accuracy.
Qwen 3.8 Max 85.9%
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 InternVL3-78B. 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 (51 entries).
- Pass Models scored 41 models we track, plus 10 other entries on its leaderboard.
- Warning Newest score The newest came 19 months ago, in April 2025, for InternVL3-78B.
- Note Frontier models None of the top 10 has a score. The strongest tested is Gemini 2.5 Pro (#104).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on huggingface.co: 42 rows for models we track and 10 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.