OpenLMM Subjective Leaderboard
Models generate open-ended answers to multimodal tasks. The board reports subjective evaluation results.
The leaderboard aggregates results from component evaluations.
- Made by
- OpenCompass
- Released
- April 2025
- 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
GPT-4.1 leads with 76.6 among models released by April 2025 GPT-4.1 leads with 76.6 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
- 76.6 GPT-4.1
- 73.8
Qwen2.5-VL-32B
- 73.7
Qwen2.5-VL-72B
- 72.5 GPT-4o (1120, detail-high)
- 71.7
Gemini 2.5 Flash
- 71.5 Claude 3.7 Sonnet
- 70.1
Gemini 2.0 Flash
- 68.9 Claude 3.5 Sonnet
- 68.9 Grok 2 Vision (1212)
- 67.6
InternVL3-78B
- 66.6 GPT-4o (0806, detail-high)
- 65.9 GPT-4o-mini (0718, detail-high)
- 65.5
Llama 4 Scout
- 64.3
InternVL3-38B
- 63.7
InternVL3-8B
- 63.2
Ovis2-34B
- 62.7
InternVL2_5-78B
- 60.4
Pixtral 12B
- 59.9
InternVL2.5-38B
- 59.7
Qwen2.5-VL-7B
- 55.7
InternVL2.5-8B
- 52.0 LLaVA-OneVision 72B
- 49.8
NVLM-D-72B
- 48.1
Llama 3.2 11B
- 48.1
DeepSeek-VL2
- 46.8 MiniCPM-o-2_6
- 44.1
LLaVA-OneVision 7B
- 42.4
Phi-4-multimodal
- 36.3
Janus-Pro-7B
- 1 GPT-4.1 76.6
- 2
Qwen2.5-VL-32B 73.8
- 3
Qwen2.5-VL-72B 73.7
- 4 GPT-4o (1120, detail-high) 72.5
- 5
Gemini 2.5 Flash Apr 2025 71.7
- 6 Claude 3.7 Sonnet 71.5
- 7
Gemini 2.0 Flash Feb 2025 70.1
- 8 Claude 3.5 Sonnet 68.9
- 9 Grok 2 Vision (1212) 68.9
- 10
InternVL3-78B 67.6
- 11 GPT-4o (0806, detail-high) 66.6
- 12 GPT-4o-mini (0718, detail-high) 65.9
- 13
Llama 4 Scout 65.5
- 14
InternVL3-38B 64.3
- 15
InternVL3-8B 63.7
- 16
Ovis2-34B 63.2
- 17
InternVL2_5-78B 62.7
- 18
Pixtral 12B 60.4
- 19
InternVL2.5-38B 59.9
- 20
Qwen2.5-VL-7B 59.7
- 21
InternVL2.5-8B 55.7
- 22 LLaVA-OneVision 72B 52.0
- 23
NVLM-D-72B 49.8
- 24
Llama 3.2 11B 48.1
- 25
DeepSeek-VL2 48.1
- 26 MiniCPM-o-2_6 46.8
- 27
LLaVA-OneVision 7B 44.1
- 28
Phi-4-multimodal 42.4
- 29
Janus-Pro-7B 36.3
Source: Official OpenLMM Subjective Leaderboard leaderboard
- 76.6 GPT-4.1
- 73.8
Qwen2.5-VL-32B
- 73.7
Qwen2.5-VL-72B
- 71.7
Gemini 2.5 Flash
- 71.5 Claude 3.7 Sonnet
- 70.1
Gemini 2.0 Flash
- 68.9 Claude 3.5 Sonnet
- 68.9 Grok 2 Vision (1212)
- 67.6
InternVL3-78B
- 65.5
Llama 4 Scout
- 64.3
InternVL3-38B
- 63.7
InternVL3-8B
- 63.2
Ovis2-34B
- 62.7
InternVL2_5-78B
- 60.4
Pixtral 12B
- 59.9
InternVL2.5-38B
- 59.7
Qwen2.5-VL-7B
- 55.7
InternVL2.5-8B
- 52.0 LLaVA-OneVision 72B
- 49.8
NVLM-D-72B
- 48.1
Llama 3.2 11B
- 48.1
DeepSeek-VL2
- 46.8 MiniCPM-o-2_6
- 44.1
LLaVA-OneVision 7B
- 42.4
Phi-4-multimodal
- 36.3
Janus-Pro-7B
- 1 GPT-4.1 76.6
- 2
Qwen2.5-VL-32B 73.8
- 3
Qwen2.5-VL-72B 73.7
- 4
Gemini 2.5 Flash Apr 2025 71.7
- 5 Claude 3.7 Sonnet 71.5
- 6
Gemini 2.0 Flash Feb 2025 70.1
- 7 Claude 3.5 Sonnet 68.9
- 8 Grok 2 Vision (1212) 68.9
- 9
InternVL3-78B 67.6
- 10
Llama 4 Scout 65.5
- 11
InternVL3-38B 64.3
- 12
InternVL3-8B 63.7
- 13
Ovis2-34B 63.2
- 14
InternVL2_5-78B 62.7
- 15
Pixtral 12B 60.4
- 16
InternVL2.5-38B 59.9
- 17
Qwen2.5-VL-7B 59.7
- 18
InternVL2.5-8B 55.7
- 19 LLaVA-OneVision 72B 52.0
- 20
NVLM-D-72B 49.8
- 21
Llama 3.2 11B 48.1
- 22
DeepSeek-VL2 48.1
- 23 MiniCPM-o-2_6 46.8
- 24
LLaVA-OneVision 7B 44.1
- 25
Phi-4-multimodal 42.4
- 26
Janus-Pro-7B 36.3
Source: Official OpenLMM Subjective Leaderboard leaderboard
| # | Entry | Score | Setting | Released | Run on | Link | Source |
|---|---|---|---|---|---|---|---|
| 1 | GPT-4.1 | 76.6 | Apr 2025 | Apr 2025 | openai.com | Official leaderboard | |
| 2 | | 73.8 | Mar 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 3 | | 73.7 | Jan 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 4 | GPT-4o (1120, detail-high) | 72.5 | Apr 2025 | openai.com | Official leaderboard | ||
| 5 | | 71.7 | Apr 2025 | Apr 2025 | deepmind.google | Official leaderboard | |
| 6 | Claude 3.7 Sonnet | 71.5 | Feb 2025 | Apr 2025 | docs.anthropic.com | Official leaderboard | |
| 7 | | 70.1 | Feb 2025 | Apr 2025 | deepmind.google | Official leaderboard | |
| 8 | Claude 3.5 Sonnet | 68.9 | Jun 2024 | Apr 2025 | docs.anthropic.com | Official leaderboard | |
| 9 | Grok 2 Vision (1212) | 68.9 | Dec 2024 | Apr 2025 | docs.x.ai | Official leaderboard | |
| 10 | | 67.6 | Apr 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 11 | GPT-4o (0806, detail-high) | 66.6 | Apr 2025 | openai.com | Official leaderboard | ||
| 12 | GPT-4o-mini (0718, detail-high) | 65.9 | Apr 2025 | openai.com | Official leaderboard | ||
| 13 | | 65.5 | Apr 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 14 | | 64.3 | Apr 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 15 | | 63.7 | Apr 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 16 | | 63.2 | Feb 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 17 | | 62.7 | Dec 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 18 | | 60.4 | Sep 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 19 | | 59.9 | Dec 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 20 | | 59.7 | Jan 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 21 | | 55.7 | Dec 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 22 | LLaVA-OneVision 72B | 52.0 | Aug 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 23 | | 49.8 | Sep 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 24 | | 48.1 | Sep 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 25 | | 48.1 | Dec 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 26 | MiniCPM-o-2_6 | 46.8 | Jan 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 27 | | 44.1 | Aug 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 28 | | 42.4 | Feb 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 29 | | 36.3 | Jan 2025 | Apr 2025 | huggingface.co | Official leaderboard |
| # | Model | Score | Setting | Released | Run on | Link | Source |
|---|---|---|---|---|---|---|---|
| 1 | GPT-4.1 | 76.6 | Apr 2025 | Apr 2025 | openai.com | Official leaderboard | |
| 2 | | 73.8 | Mar 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 3 | | 73.7 | Jan 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 4 | | 71.7 | Apr 2025 | Apr 2025 | deepmind.google | Official leaderboard | |
| 5 | Claude 3.7 Sonnet | 71.5 | Feb 2025 | Apr 2025 | docs.anthropic.com | Official leaderboard | |
| 6 | | 70.1 | Feb 2025 | Apr 2025 | deepmind.google | Official leaderboard | |
| 7 | Claude 3.5 Sonnet | 68.9 | Jun 2024 | Apr 2025 | docs.anthropic.com | Official leaderboard | |
| 8 | Grok 2 Vision (1212) | 68.9 | Dec 2024 | Apr 2025 | docs.x.ai | Official leaderboard | |
| 9 | | 67.6 | Apr 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 10 | | 65.5 | Apr 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 11 | | 64.3 | Apr 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 12 | | 63.7 | Apr 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 13 | | 63.2 | Feb 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 14 | | 62.7 | Dec 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 15 | | 60.4 | Sep 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 16 | | 59.9 | Dec 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 17 | | 59.7 | Jan 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 18 | | 55.7 | Dec 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 19 | LLaVA-OneVision 72B | 52.0 | Aug 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 20 | | 49.8 | Sep 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 21 | | 48.1 | Sep 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 22 | | 48.1 | Dec 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 23 | MiniCPM-o-2_6 | 46.8 | Jan 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 24 | | 44.1 | Aug 2024 | Apr 2025 | huggingface.co | Official leaderboard | |
| 25 | | 42.4 | Feb 2025 | Apr 2025 | huggingface.co | Official leaderboard | |
| 26 | | 36.3 | Jan 2025 | Apr 2025 | huggingface.co | Official leaderboard |
Rankings by part
We read 7 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Creation_MMBench reward, Creation_MMBench VFS and MIA-Bench. Qwen2.5-VL-32B leads 2 of them, and 4 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
Creation_MMBench reward
26 models we track have a score here.
- 1
Qwen2.5-VL-32B ↑1 6.97
- 2
Gemini 2.5 Flash Apr 2025 ↑2 6.61
- 3 GPT-4.1 ↓2 5.61
- 4
Qwen2.5-VL-72B ↓1 -5.82
- 5
InternVL3-78B ↑4 -6.87
- 6 Grok 2 Vision (1212) ↑2 -8.86
- 7
Gemini 2.0 Flash Feb 2025 ↓1 -8.90
- 8
InternVL3-38B ↑3 -8.99
- 9 Claude 3.7 Sonnet ↓4 -11.28
- 10
InternVL3-8B ↑2 -14.00
Creation_MMBench VFS
26 models we track have a score here.
- 1 GPT-4.1 8.96
- 2
Gemini 2.5 Flash Apr 2025 ↑2 8.74
- 3
Gemini 2.0 Flash Feb 2025 ↑3 8.42
- 4
Qwen2.5-VL-72B ↓1 8.33
- 5
Qwen2.5-VL-32B ↓3 8.29
- 6
InternVL3-78B ↑3 8.25
- 7
InternVL3-38B ↑4 8.18
- 8
Ovis2-34B ↑5 8.17
- 9 Claude 3.7 Sonnet ↓4 8.03
- 10
Llama 4 Scout 7.97
MIA-Bench
26 models we track have a score here.
- 1 GPT-4.1 94.5
- 2 Grok 2 Vision (1212) ↑6 93.6
- 3
Qwen2.5-VL-72B 93.2
- 4
Gemini 2.0 Flash Feb 2025 ↑2 92.9
- 5
InternVL3-78B ↑4 92.7
- 6
Gemini 2.5 Flash Apr 2025 ↓2 92.2
- 7
InternVL3-38B ↑4 92.0
- 8
InternVL2_5-78B ↑6 92.0
- 9
Llama 4 Scout ↑1 91.6
- 10 LLaVA-OneVision 72B ↑9 91.5
MM-IFEval
26 models we track have a score here.
- 1
Qwen2.5-VL-72B ↑2 73.7
- 2 GPT-4.1 ↓1 72.8
- 3
Gemini 2.5 Flash Apr 2025 ↑1 71.3
- 4
Llama 4 Scout ↑6 68.5
- 5
Gemini 2.0 Flash Feb 2025 ↑1 67.2
- 6 Claude 3.7 Sonnet ↓1 64.4
- 7
InternVL3-78B ↑2 63.9
- 8 Claude 3.5 Sonnet ↓1 63.7
- 9 Grok 2 Vision (1212) ↓1 61.5
- 10
Ovis2-34B ↑3 59.0
MMAlignBench
26 models we track have a score here.
- 1
Qwen2.5-VL-32B ↑1 75.0
- 2 Claude 3.7 Sonnet ↑3 60.3
- 3 GPT-4.1 ↓2 59.9
- 4 Grok 2 Vision (1212) ↑4 53.2
- 5
Qwen2.5-VL-72B ↓2 50.0
- 6
Gemini 2.0 Flash Feb 2025 40.3
- 7 Claude 3.5 Sonnet 34.9
- 8
Gemini 2.5 Flash Apr 2025 ↓4 32.1
- 9
Pixtral 12B ↑6 21.2
- 10
Llama 4 Scout 14.3
MMVet
26 models we track have a score here.
- 1
InternVL3-8B ↑11 83.2
- 2
InternVL3-78B ↑7 83.2
- 3
Gemini 2.5 Flash Apr 2025 ↑1 81.3
- 4 GPT-4.1 ↓3 80.1
- 5
Ovis2-34B ↑8 77.4
- 6
Gemini 2.0 Flash Feb 2025 77.0
- 7
Qwen2.5-VL-72B ↓4 76.8
- 8
InternVL3-38B ↑3 75.1
- 9
Qwen2.5-VL-32B ↓7 73.1
- 10
InternVL2_5-78B ↑4 72.7
WildVision
26 models we track have a score here.
- 1 Claude 3.7 Sonnet ↑4 60.6
- 2
Qwen2.5-VL-32B 58.7
- 3 GPT-4.1 ↓2 58.7
- 4 Claude 3.5 Sonnet ↑3 53.2
- 5
Qwen2.5-VL-72B ↓2 52.3
- 6
InternVL3-38B ↑5 45.2
- 7
InternVL2_5-78B ↑7 42.4
- 8
Ovis2-34B ↑5 40.9
- 9 Grok 2 Vision (1212) ↓1 40.0
- 10
InternVL3-78B ↓1 39.6
2 models have held number one on the official leaderboard since its release in April 2025
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official OpenLMM Subjective 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 OpenLMM Subjective 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 Gemini 2.5 Flash. 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 (29 entries).
- Pass Models scored 26 models we track, plus 3 other entries on its leaderboard.
- Warning Newest score The newest came 19 months ago, in April 2025, for Gemini 2.5 Flash.
- 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: 26 rows for models we track and 3 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.