{ "leaderboardData": [ { "info": { "name": "Human Expert (High)", "size": "-", "date": "2024-01-31", "type": "human_expert" }, "validation": { "overall": "88.6", "artDesign": "89.2", "business": "90.7", "science": "90.0", "healthMedicine": "87.3", "humanSocialSci": "89.2", "techEng": "86.2" }, "pro": { "overall": "85.4", "vision": "85.4", "original": "85.4" } }, { "info": { "name": "Human Expert (Medium)", "size": "-", "date": "2024-01-31", "type": "human_expert" }, "validation": { "overall": "82.6", "artDesign": "84.2", "business": "86.0", "science": "84.7", "healthMedicine": "78.8", "humanSocialSci": "85.0", "techEng": "79.1" }, "pro": { "overall": "80.8", "vision": "80.8", "original": "80.8" } }, { "info": { "name": "Human Expert (Low)", "size": "-", "date": "2024-01-31", "type": "human_expert" }, "validation": { "overall": "76.2", "artDesign": "80.8", "business": "78.0", "science": "78.0", "healthMedicine": "73.3", "humanSocialSci": "74.2", "techEng": "74.3" }, "pro": { "overall": "73.0", "vision": "73.0", "original": "73.0" } }, { "info": { "name": "Muse Spark Thinking", "size": "-", "date": "2026-04-08", "type": "proprietary", "link": "https://ai.meta.com/blog/introducing-muse-spark-msl/" }, "pro": { "overall": "80.4", "source": "author" } }, { "info": { "name": "Claude Opus 4.6 w/o tools", "size": "-", "date": "2026-02-05", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-opus-4-6" }, "pro": { "overall": "73.9", "source": "author" } }, { "info": { "name": "Claude Opus 4.6 w/ tools", "size": "-", "date": "2026-02-05", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-opus-4-6" }, "pro": { "overall": "77.3", "source": "author" } }, { "info": { "name": "Claude Sonnet 4.6 w/o tools", "size": "-", "date": "2026-02-17", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-sonnet-4-6" }, "pro": { "overall": "74.5", "source": "author" } }, { "info": { "name": "Claude Sonnet 4.6 w/ tools", "size": "-", "date": "2026-02-17", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-sonnet-4-6" }, "pro": { "overall": "75.6", "source": "author" } }, { "info": { "name": "GPT-5.4 Thinking w/o tools", "size": "-", "date": "2026-03-05", "type": "proprietary", "link": "https://openai.com/index/introducing-gpt-5-4/" }, "pro": { "overall": "81.2", "source": "author" } }, { "info": { "name": "GPT-5.4 Thinking w/ tools", "size": "-", "date": "2026-03-05", "type": "proprietary", "link": "https://openai.com/index/introducing-gpt-5-4/" }, "pro": { "overall": "82.1", "source": "author" } }, { "info": { "name": "GPT-5.2 Thinking w/o tools", "size": "-", "date": "2025-12-11", "type": "proprietary", "link": "https://openai.com/index/introducing-gpt-5-2/" }, "pro": { "overall": "79.5", "source": "author" } }, { "info": { "name": "GPT-5.2 Thinking w/o Python", "size": "-", "date": "2025-12-11", "type": "proprietary", "link": "https://openai.com/index/introducing-gpt-5-2/" }, "pro": { "overall": "80.4", "source": "author" } }, { "info": { "name": "GPT-5.1 Thinking", "size": "-", "date": "2025-11-13", "type": "proprietary", "link": "https://openai.com/index/gpt-5-1-for-developers/" }, "pro": { "overall": "79.0", "source": "author" } }, { "info": { "name": "Gemini 3.1 Pro Thinking (High)", "size": "-", "date": "2026-02-19", "type": "proprietary", "link": "https://deepmind.google/models/gemini/pro/" }, "pro": { "overall": "80.5", "source": "author" } }, { "info": { "name": "Gemma 4 31B", "size": "31B", "date": "2026-04-02", "type": "open_source", "link": "https://deepmind.google/models/gemma/gemma-4/" }, "pro": { "overall": "76.9", "source": "author" } }, { "info": { "name": "Gemma 4 26B A4B", "size": "26B", "date": "2026-04-02", "type": "open_source", "link": "https://deepmind.google/models/gemma/gemma-4/" }, "pro": { "overall": "73.8", "source": "author" } }, { "info": { "name": "Gemma 4 E4B", "size": "4B", "date": "2026-04-02", "type": "open_source", "link": "https://deepmind.google/models/gemma/gemma-4/" }, "pro": { "overall": "52.6", "source": "author" } }, { "info": { "name": "Gemma 4 E2B", "size": "2B", "date": "2026-04-02", "type": "open_source", "link": "https://deepmind.google/models/gemma/gemma-4/" }, "pro": { "overall": "44.2", "source": "author" } }, { "info": { "name": "GPT-5.1", "size": "-", "date": "2025-11-13", "type": "proprietary", "link": "https://openai.com/index/gpt-5-1-for-developers/" }, "validation": { "overall": "85.4", "source": "author" }, "pro": { "overall": "76.0", "source": "author" } }, { "info": { "name": "GPT-5 w/ thinking", "size": "-", "date": "2025-08-07", "type": "proprietary", "link": "https://openai.com/gpt-5/" }, "validation": { "overall": "84.2", "source": "author" }, "pro": { "overall": "78.4", "source": "author" } }, { "info": { "name": "GPT-5 w/o thinking", "size": "-", "date": "2025-08-07", "type": "proprietary", "link": "https://openai.com/gpt-5/" }, "validation": { "overall": "74.4", "source": "author" }, "pro": { "overall": "62.7", "source": "author" } }, { "info": { "name": "o3", "size": "-", "date": "2025-04-16", "type": "proprietary", "link": "https://openai.com/index/introducing-o3-and-o4-mini/" }, "validation": { "overall": "82.9", "source": "author" }, "pro": { "overall": "76.4", "source": "author" } }, { "info": { "name": "o4-mini", "size": "-", "date": "2025-04-16", "type": "proprietary", "link": "https://openai.com/index/introducing-o3-and-o4-mini/" }, "validation": { "overall": "81.6", "source": "author" } }, { "info": { "name": "o1", "size": "-", "date": "2024-09-12", "type": "proprietary", "link": "https://openai.com/index/introducing-openai-o1-preview/" }, "validation": { "overall": "78.2", "source": "author" } }, { "info": { "name": "GPT‑4.5", "size": "-", "date": "2025-02-28", "type": "proprietary", "link": "GPT‑4.5" }, "validation": { "overall": "74.4", "source": "author" } }, { "info": { "name": "GPT-4o (0513)", "size": "-", "date": "2024-05-13", "type": "proprietary", "link": "https://openai.com/index/hello-gpt-4o/" }, "validation": { "overall": "69.1", "source": "author" }, "pro": { "overall": "51.9", "vision": "49.7", "original": "54.0" } }, { "info": { "name": "GPT-4o (1120)", "size": "-", "date": "2024-11-20", "type": "proprietary", "link": "https://openai.com/index/hello-gpt-4o/" }, "validation": { "overall": "70.7", "source": "author" } }, { "info": { "name": "Claude Opus 4.1", "size": "-", "date": "2025-08-05", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-opus-4-1" }, "validation": { "overall": "77.1", "source": "author" } }, { "info": { "name": "Claude Opus 4.5", "size": "-", "date": "2025-11-24", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-opus-4-5" }, "validation": { "overall": "80.7", "source": "author" }, "pro": { "overall": "73.9", "source": "author" } }, { "info": { "name": "Claude Sonnet 4.5", "size": "-", "date": "2025-09-29", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-sonnet-4-5" }, "validation": { "overall": "77.8", "source": "author" }, "pro": { "overall": "68.9", "source": "author" } }, { "info": { "name": "Claude Sonnet 4.5 w/o tools", "size": "-", "date": "2025-09-29", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-sonnet-4-5" }, "pro": { "overall": "63.4", "source": "author" } }, { "info": { "name": "Claude Sonnet 4", "size": "-", "date": "2025-05-23", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-4" }, "validation": { "overall": "74.4", "source": "author" } }, { "info": { "name": "Claude Opus 4", "size": "-", "date": "2025-05-23", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-4" }, "validation": { "overall": "76.5", "source": "author" } }, { "info": { "name": "Claude 3.5 Sonnet", "size": "-", "date": "2024-06-20", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-3-5-sonnet" }, "validation": { "overall": "68.3" }, "pro": { "overall": "51.5", "vision": "48.0", "original": "55.0" } }, { "info": { "name": "Claude 3.7 Sonnet", "size": "-", "date": "2025-02-24", "type": "proprietary", "link": "https://www.anthropic.com/claude/sonnet" }, "validation": { "overall": "75.0", "source": "author" } }, { "info": { "name": "Grok 3 Beta", "size": "-", "date": "2025-02-19", "type": "proprietary", "link": "https://x.ai/news/grok-3" }, "validation": { "overall": "78.0", "source": "author" } }, { "info": { "name": "Gemini 3.0 Pro", "size": "-", "date": "2025-11-18", "type": "proprietary", "link": "https://blog.google/products/gemini/gemini-3/#gemini-3" }, "pro": { "overall": "81.0", "source": "author" } }, { "info": { "name": "Gemini 2.0 Flash-Lite", "size": "-", "date": "2025-02-05", "type": "proprietary", "link": "https://deepmind.google/technologies/gemini/" }, "validation": { "overall": "68.0", "source": "author" } }, { "info": { "name": "Gemini 2.0 Flash", "size": "-", "date": "2025-02-05", "type": "proprietary", "link": "https://deepmind.google/technologies/gemini/" }, "validation": { "overall": "71.7", "source": "author" } }, { "info": { "name": "Gemini 2.0 Pro", "size": "-", "date": "2025-02-05", "type": "proprietary", "link": "https://deepmind.google/technologies/gemini/" }, "validation": { "overall": "72.7", "source": "author" } }, { "info": { "name": "Gemini 2.5 Pro Deep-Think", "size": "-", "date": "2025-05-20", "type": "proprietary", "link": "https://deepmind.google/technologies/gemini/pro/" }, "validation": { "overall": "84.0", "source": "author" } }, { "info": { "name": "Gemini 2.5 Pro 05-06", "size": "-", "date": "2025-05-06", "type": "proprietary", "link": "https://deepmind.google/technologies/gemini/pro/" }, "validation": { "overall": "79.6", "source": "author" }, "pro": { "overall": "68.0", "source": "author" } }, { "info": { "name": "Gemini 2.5 Flash 05-20", "size": "-", "date": "2025-05-20", "type": "proprietary", "link": "https://deepmind.google/models/gemini/flash/" }, "validation": { "overall": "79.7", "source": "author" } }, { "info": { "name": "Gemini 1.5 Pro (0801)", "size": "-", "date": "2024-08-01", "type": "proprietary", "link": "https://deepmind.google/technologies/gemini/pro/" }, "validation": { "overall": "65.8" }, "pro": { "overall": "46.9", "vision": "44.4", "original": "49.4" } }, { "info": { "name": "Gemini 1.5 Pro (0523)", "size": "-", "date": "2024-05-23", "type": "proprietary", "link": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v1_5_report.pdf" }, "validation": { "overall": "62.2", "source": "author" }, "pro": { "overall": "43.5", "vision": "40.5", "original": "46.5" } }, { "info": { "name": "dots.vlm1", "size": "37B", "date": "2025-08-05", "type": "open_source", "link": "https://github.com/rednote-hilab/dots.vlm1" }, "validation": { "overall": "80.1", "source": "author" }, "pro": { "overall": "70.1", "source": "author" } }, { "info": { "name": "Qwen3-VL 235B-A22B", "size": "22B", "date": "2025-09-22", "type": "open_source", "link": "https://qwen.ai/blog?id=99f0335c4ad9ff6153e517418d48535ab6d8afef&from=research.latest-advancements-list" }, "validation": { "overall": "78.7", "source": "author" }, "pro": { "overall": "68.1", "source": "author" } }, { "info": { "name": "Qianfan-VL-3B", "size": "3B", "date": "2025-09-16", "type": "open_source", "link": "https://baidubce.github.io/Qianfan-VL/index_en.html" }, "validation": { "overall": "46.4", "source": "author" } }, { "info": { "name": "Qianfan-VL-8B", "size": "8B", "date": "2025-09-16", "type": "open_source", "link": "https://baidubce.github.io/Qianfan-VL/index_en.html" }, "validation": { "overall": "47.1", "source": "author" } }, { "info": { "name": "Qianfan-VL-70B", "size": "70B", "date": "2025-09-16", "type": "open_source", "link": "https://baidubce.github.io/Qianfan-VL/index_en.html" }, "validation": { "overall": "58.3", "source": "author" } }, { "info": { "name": "Intern-S1", "size": "22B", "date": "2025-08-04", "type": "open_source", "link": "https://github.com/InternLM/Intern-S1/" }, "validation": { "overall": "77.7", "source": "author" } }, { "info": { "name": "InternVL2-Pro", "size": "-", "date": "2024-07-04", "type": "proprietary", "link": "https://internvl.github.io/blog/2024-07-02-InternVL-2.0/" }, "validation": { "overall": "62.0", "source": "author" }, "test": { "overall": "55.7", "artDesign": "67.4", "business": "52.4", "science": "48.3", "healthMedicine": "63.2", "humanSocialSci": "78.0", "techEng": "46.6" } }, { "info": { "name": "Gemini 1.0 Ultra", "size": "-", "date": "2023-12-06", "type": "proprietary", "link": "https://deepmind.google/technologies/gemini/#introduction" }, "validation": { "overall": "59.4", "artDesign": "70.0", "business": "56.7", "science": "48.0", "healthMedicine": "67.3", "humanSocialSci": "78.3", "techEng": "47.1", "source": "author" } }, { "info": { "name": "Claude 3 Opus", "size": "-", "date": "2024-03-05", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-3-family" }, "validation": { "overall": "59.4", "artDesign": "67.5", "business": "67.2", "science": "48.9", "healthMedicine": "61.1", "humanSocialSci": "70.0", "techEng": "50.6", "source": "author" } }, { "info": { "name": "GPT-4V(ision) (Playground)", "size": "-", "date": "2023-11-27", "type": "proprietary", "link": "https://openai.com/contributions/gpt-4v" }, "validation": { "overall": "56.8", "artDesign": "65.8", "business": "59.3", "science": "54.7", "healthMedicine": "64.7", "humanSocialSci": "72.5", "techEng": "36.7" }, "test": { "overall": "56.1", "artDesign": "65.5", "business": "65.5", "science": "49.2", "healthMedicine": "63.3", "humanSocialSci": "76.3", "techEng": "41.5" } }, { "info": { "name": "KiQ-v0", "size": "-", "date": "2025-08-31", "type": "proprietary", "link": "https://uark-cviu.github.io/projects/papers/BiMax.pdf" }, "validation": { "overall": "72.6", "artDesign": "75.8", "business": "81.3", "science": "64.0", "healthMedicine": "74.0", "humanSocialSci": "80.8", "techEng": "64.8" }, "test": { "overall": "65.9", "artDesign": "64.6", "business": "78.3", "science": "62.7", "healthMedicine": "68.3", "humanSocialSci": "72.1", "techEng": "59.3" } }, { "info": { "name": "Reka Core", "size": "-", "date": "2024-04-23", "type": "proprietary", "link": "https://publications.reka.ai/reka-core-tech-report.pdf" }, "validation": { "overall": "56.3", "artDesign": "75.9", "business": "47.3", "science": "49.3", "healthMedicine": "58.0", "humanSocialSci": "75.0", "techEng": "44.2", "source": "author" } }, { "info": { "name": "LLaVA-Critic-R1", "size": "7B", "date": "2025-08-30", "type": "open_source", "link": "https://arxiv.org/pdf/2509.00676" }, "validation": { "overall": "69.7", "source": "author" } }, { "info": { "name": "LLaVA-Critic-R1+", "size": "7B", "date": "2025-08-30", "type": "open_source", "link": "https://arxiv.org/pdf/2509.00676" }, "validation": { "overall": "71.9", "source": "author" } }, { "info": { "name": "Gemini 1.5 Flash", "size": "-", "date": "2024-05-23", "type": "proprietary", "link": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v1_5_report.pdf" }, "validation": { "overall": "56.1", "source": "author" } }, { "info": { "name": "SenseChat-Vision-0423-Preview", "size": "-", "date": "2024-04-23", "type": "proprietary", "link": "https://platform.sensenova.cn/doc?path=/model/mllm.md" }, "validation": { "overall": "54.6", "artDesign": "66.7", "business": "54.0", "science": "45.3", "healthMedicine": "53.3", "humanSocialSci": "75.0", "techEng": "43.8", "source": "author" }, "test": { "overall": "50.3", "artDesign": "62.8", "business": "44.1", "science": "42.3", "healthMedicine": "55.7", "humanSocialSci": "74.7", "techEng": "43.5", "source": "author" } }, { "info": { "name": "Reka Flash", "size": "-", "date": "2024-04-23", "type": "proprietary", "link": "https://reka.ai/reka-flash-an-efficient-and-capable-multimodal-language-model/" }, "validation": { "overall": "53.3", "artDesign": "61.7", "business": "42.7", "science": "47.3", "healthMedicine": "59.3", "humanSocialSci": "74.2", "techEng": "44.3", "source": "author" } }, { "info": { "name": "Claude 3 Sonnet", "size": "-", "date": "2024-03-05", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-3-family" }, "validation": { "overall": "53.1", "artDesign": "61.7", "business": "58.2", "science": "37.1", "healthMedicine": "57.1", "humanSocialSci": "68.7", "techEng": "45.0", "source": "author" } }, { "info": { "name": "HPT Pro", "size": "-", "date": "2024-03-16", "type": "proprietary", "link": "https://huggingface.co/HyperGAI" }, "validation": { "overall": "52.0", "artDesign": "66.7", "business": "43.3", "science": "42.7", "healthMedicine": "50.7", "humanSocialSci": "72.5", "techEng": "43.8", "source": "author" } }, { "info": { "name": "NVILA", "size": "15B", "date": "2024-11-20", "type": "open_source", "link": "https://github.com/NVlabs/VILA" }, "validation": { "overall": "58.7", "artDesign": "65.8", "business": "62.7", "science": "50.0", "healthMedicine": "61.3", "humanSocialSci": "74.2", "techEng": "47.1", "source": "author" }, "test": { "overall": "51.8", "artDesign": "63.2", "business": "54.6", "science": "42.9", "healthMedicine": "57.0", "humanSocialSci": "73.7", "techEng": "42.7", "source": "author" }, "pro": { "overall": "33.7", "vision": "28.6", "original": "38.7", "source": "author" } }, { "info": { "name": "InternVL-Chat-V1.2", "size": "-", "date": "2024-02-22", "type": "open_source", "link": "https://github.com/OpenGVLab/InternVL/" }, "validation": { "overall": "51.6", "artDesign": "62.5", "business": "40.7", "science": "39.3", "healthMedicine": "58.7", "humanSocialSci": "70.0", "techEng": "46.2", "source": "author" }, "test": { "overall": "46.2", "artDesign": "62.4", "business": "37.6", "science": "37.9", "healthMedicine": "49.7", "humanSocialSci": "70.1", "techEng": "40.8", "source": "author" } }, { "info": { "name": "Qwen-VL-MAX", "size": "-", "date": "2024-01-27", "type": "proprietary", "link": "https://github.com/QwenLM/Qwen-VL/tree/master" }, "validation": { "overall": "51.4", "artDesign": "72.5", "business": "43.3", "science": "40.0", "healthMedicine": "58.0", "humanSocialSci": "69.2", "techEng": "38.6", "source": "author" }, "test": { "overall": "46.8", "artDesign": "64.1", "business": "39.8", "science": "36.3", "healthMedicine": "52.5", "humanSocialSci": "70.3", "techEng": "40.7", "source": "author" } }, { "info": { "name": "Skywork-VL", "size": "-", "date": "2024-05-30", "type": "proprietary", "link": "https://www.kunlun.com/en/" }, "validation": { "overall": "51.4", "artDesign": "66.7", "business": "41.3", "science": "38.7", "healthMedicine": "55.3", "humanSocialSci": "68.3", "techEng": "46.7", "source": "author" }, "test": { "overall": "46.2", "artDesign": "61.3", "business": "39.6", "science": "36.6", "healthMedicine": "50.8", "humanSocialSci": "71.6", "techEng": "40.2", "source": "author" } }, { "info": { "name": "LLaVA-NEXT-34B", "size": "34B", "date": "2024-01-30", "type": "open_source", "link": "https://llava-vl.github.io/blog/2024-01-30-llava-1-6/" }, "validation": { "overall": "48.1", "source": "author" }, "test": { "overall": "44.7", "artDesign": "58.5", "business": "39.9", "science": "36.0", "healthMedicine": "51.2", "humanSocialSci": "70.2", "techEng": "36.3", "source": "author" }, "pro": { "overall": "23.8", "vision": "17.2", "original": "30.3" } }, { "info": { "name": "Claude 3 Haiku", "size": "-", "date": "2024-03-05", "type": "proprietary", "link": "https://www.anthropic.com/news/claude-3-family" }, "validation": { "overall": "50.2", "artDesign": "60.8", "business": "52.5", "science": "37.1", "healthMedicine": "52.3", "humanSocialSci": "66.0", "techEng": "41.5", "source": "author" } }, { "info": { "name": "Idefics3-8B-Llama3", "size": "8B", "date": "2024-08-22", "type": "open_source", "link": "https://huggingface.co/HuggingFaceM4/Idefics3-8B-Llama3" }, "validation": { "overall": "46.6", "source": "author" }, "pro": { "overall": "22.9", "vision": "15.6", "original": "30.1" } }, { "info": { "name": "Adept Fuyu-Heavy", "size": "-", "date": "2024-01-31", "type": "proprietary", "link": "https://www.adept.ai/blog/adept-fuyu-heavy" }, "validation": { "overall": "48.3", "artDesign": "53.4", "business": "46.3", "science": "33.7", "healthMedicine": "51.3", "humanSocialSci": "72.2", "techEng": "44.0", "source": "author" } }, { "info": { "name": "Gemini 1.0 Pro", "size": "-", "date": "2023-12-06", "type": "proprietary", "link": "https://deepmind.google/technologies/gemini/#introduction" }, "validation": { "overall": "47.9", "source": "author" } }, { "info": { "name": "Marco-VL-Plus", "size": "-", "date": "2024-03-09", "type": "proprietary" }, "validation": { "overall": "46.2", "artDesign": "60.8", "business": "37.3", "science": "35.3", "healthMedicine": "48.7", "humanSocialSci": "69.2", "techEng": "37.1", "source": "author" }, "test": { "overall": "44.3", "artDesign": "57.4", "business": "34.7", "science": "38.5", "healthMedicine": "48.7", "humanSocialSci": "72.2", "techEng": "36.7", "source": "author" } }, { "info": { "name": "Yi-VL-34B", "size": "34B", "date": "2023-12-31", "type": "open_source", "link": "https://huggingface.co/01-ai/Yi-VL-34B" }, "validation": { "overall": "45.9", "artDesign": "59.2", "business": "36.0", "science": "33.3", "healthMedicine": "51.3", "humanSocialSci": "62.5", "techEng": "41.0", "source": "author" }, "test": { "overall": "41.5", "artDesign": "56.0", "business": "33.3", "science": "32.9", "healthMedicine": "45.9", "humanSocialSci": "66.5", "techEng": "36.0", "source": "author" } }, { "info": { "name": "Qwen-VL-PLUS", "size": "-", "date": "2023-12-15", "type": "proprietary", "link": "https://github.com/QwenLM/Qwen-VL?tab=readme-ov-file#qwen-vl-plus" }, "validation": { "overall": "45.2", "artDesign": "60.0", "business": "35.3", "science": "37.3", "healthMedicine": "46.7", "humanSocialSci": "65.8", "techEng": "36.7", "source": "author" }, "test": { "overall": "40.8", "artDesign": "59.8", "business": "34.5", "science": "32.8", "healthMedicine": "43.7", "humanSocialSci": "65.4", "techEng": "32.9", "source": "author" } }, { "info": { "name": "HPT Air", "size": "-", "date": "2024-03-16", "type": "open_source", "link": "https://huggingface.co/HyperGAI" }, "validation": { "overall": "44.0", "artDesign": "63.3", "business": "31.3", "science": "34.7", "healthMedicine": "45.3", "humanSocialSci": "59.2", "techEng": "42.9", "source": "author" } }, { "info": { "name": "Bunny-8B", "size": "8B", "date": "2024-06-25", "type": "open_source", "link": "https://huggingface.co/BAAI/Bunny-v1_1-Llama-3-8B-V" }, "validation": { "overall": "43.3", "artDesign": "62.5", "business": "32.7", "science": "30.7", "healthMedicine": "43.3", "humanSocialSci": "60.8", "techEng": "39.0", "source": "author" }, "test": { "overall": "39.0", "artDesign": "56.4", "business": "30.6", "science": "30.9", "healthMedicine": "43.3", "humanSocialSci": "60.7", "techEng": "33.2", "source": "author" } }, { "info": { "name": "InternLM-XComposer2-VL", "size": "-", "date": "2024-01-22", "type": "open_source", "link": "https://github.com/InternLM/InternLM-XComposer" }, "validation": { "overall": "43.0", "artDesign": "60.0", "business": "34.0", "science": "34.7", "healthMedicine": "46.0", "humanSocialSci": "62.5", "techEng": "32.4", "source": "author" }, "test": { "overall": "38.2", "artDesign": "56.8", "business": "32.8", "science": "30.1", "healthMedicine": "39.8", "humanSocialSci": "60.7", "techEng": "31.8", "source": "author" } }, { "info": { "name": "Reka Edge", "size": "-", "date": "2024-04-23", "type": "proprietary", "link": "https://www.reka.ai/" }, "validation": { "overall": "42.8", "artDesign": "52.5", "business": "36.0", "science": "42.7", "healthMedicine": "41.3", "humanSocialSci": "59.2", "techEng": "33.8", "source": "author" } }, { "info": { "name": "Bunny-4B", "size": "4B", "date": "2024-06-25", "type": "open_source", "link": "https://huggingface.co/BAAI/Bunny-v1_1-4B" }, "validation": { "overall": "41.4", "artDesign": "60.0", "business": "36.7", "science": "29.3", "healthMedicine": "40.7", "humanSocialSci": "61.7", "techEng": "31.9", "source": "author" }, "test": { "overall": "38.4", "artDesign": "54.2", "business": "31.9", "science": "32.4", "healthMedicine": "40.4", "humanSocialSci": "60.2", "techEng": "31.7", "source": "author" } }, { "info": { "name": "Weitu-VL-1.0-15B", "size": "15B", "date": "2024-04-01", "type": "proprietary" }, "test": { "overall": "38.4", "artDesign": "56.5", "business": "30.5", "science": "31.1", "healthMedicine": "38.4", "humanSocialSci": "59.0", "techEng": "34.2", "source": "author" } }, { "info": { "name": "Marco-VL", "size": "-", "date": "2024-01-16", "type": "proprietary" }, "validation": { "overall": "41.2", "artDesign": "57.5", "business": "30.0", "science": "28.0", "healthMedicine": "45.3", "humanSocialSci": "65.8", "techEng": "32.4", "source": "author" }, "test": { "overall": "40.4", "artDesign": "56.4", "business": "31.0", "science": "31.0", "healthMedicine": "46.9", "humanSocialSci": "66.5", "techEng": "33.8", "source": "author" } }, { "info": { "name": "OmniLMM-12B", "size": "12B", "date": "2024-02-07", "type": "open_source", "link": "https://huggingface.co/openbmb/OmniLMM-12B" }, "validation": { "overall": "41.1", "artDesign": "58.3", "business": "34.0", "science": "27.3", "healthMedicine": "44.0", "humanSocialSci": "62.5", "techEng": "31.9", "source": "author" }, "test": { "overall": "37.8", "artDesign": "55.0", "business": "30.7", "science": "28.0", "healthMedicine": "40.4", "humanSocialSci": "57.6", "techEng": "34.4", "source": "author" } }, { "info": { "name": "InfiMM-Zephyr-7B", "size": "7B", "date": "2024-01-18", "type": "open_source", "link": "https://huggingface.co/Infi-MM/infimm-zephyr" }, "validation": { "overall": "39.4", "artDesign": "55.8", "business": "28.0", "science": "33.3", "healthMedicine": "42.7", "humanSocialSci": "59.2", "techEng": "29.0", "source": "author" }, "test": { "overall": "35.5", "artDesign": "50.0", "business": "29.6", "science": "28.2", "healthMedicine": "37.5", "humanSocialSci": "54.6", "techEng": "31.1", "source": "author" } }, { "info": { "name": "Yi-VL-6B", "size": "6B", "date": "2024-01-23", "type": "open_source", "link": "https://huggingface.co/01-ai/Yi-VL-6B" }, "validation": { "overall": "39.1", "artDesign": "52.5", "business": "30.7", "science": "31.3", "healthMedicine": "38.0", "humanSocialSci": "53.3", "techEng": "35.7", "source": "author" }, "test": { "overall": "37.8", "artDesign": "53.3", "business": "30.3", "science": "30.0", "healthMedicine": "39.3", "humanSocialSci": "58.4", "techEng": "34.1", "source": "author" } }, { "info": { "name": "InternVL-Chat-V1.1", "size": "-", "date": "2024-01-24", "type": "open_source", "link": "https://github.com/OpenGVLab/InternVL/" }, "validation": { "overall": "39.1", "artDesign": "56.7", "business": "34.7", "science": "31.3", "healthMedicine": "39.3", "humanSocialSci": "57.5", "techEng": "27.1", "source": "author" }, "test": { "overall": "35.3", "artDesign": "53.6", "business": "31.7", "science": "28.2", "healthMedicine": "36.5", "humanSocialSci": "56.3", "techEng": "28.0", "source": "author" } }, { "info": { "name": "Math-LLaVA-13B", "size": "13B", "date": "2024-06-25", "type": "open_source", "link": "https://arxiv.org/abs/2406.17294" }, "validation": { "overall": "38.3", "artDesign": "53.3", "business": "24.7", "science": "30.7", "healthMedicine": "38.7", "humanSocialSci": "58.3", "techEng": "33.3", "source": "author" }, "test": { "overall": "34.6", "artDesign": "50.4", "business": "27.8", "science": "28.2", "healthMedicine": "35.4", "humanSocialSci": "55.5", "techEng": "29.3", "source": "author" } }, { "info": { "name": "Bunny-3B", "size": "3B", "date": "2024-02-13", "type": "open_source", "link": "https://github.com/BAAI-DCAI/Bunny" }, "validation": { "overall": "38.2", "artDesign": "49.2", "business": "30.7", "science": "30.7", "healthMedicine": "40.7", "humanSocialSci": "45.0", "techEng": "37.1", "source": "author" }, "test": { "overall": "33.0", "artDesign": "44.2", "business": "29.5", "science": "26.8", "healthMedicine": "34.5", "humanSocialSci": "50.6", "techEng": "28.7", "source": "author" } }, { "info": { "name": "SVIT", "size": "-", "date": "2023-12-26", "type": "open_source", "link": "https://arxiv.org/abs/2307.04087" }, "validation": { "overall": "38.0", "artDesign": "52.5", "business": "27.3", "science": "28.0", "healthMedicine": "42.0", "humanSocialSci": "51.7", "techEng": "33.8", "source": "author" }, "test": { "overall": "34.1", "artDesign": "48.8", "business": "28.0", "science": "26.8", "healthMedicine": "35.5", "humanSocialSci": "50.9", "techEng": "30.7", "source": "author" } }, { "info": { "name": "MiniCPM-V", "size": "-", "date": "2024-02-01", "type": "open_source", "link": "https://huggingface.co/openbmb/MiniCPM-V" }, "validation": { "overall": "37.2", "artDesign": "55.8", "business": "33.3", "science": "28.0", "healthMedicine": "32.7", "humanSocialSci": "58.3", "techEng": "27.1", "source": "author" } }, { "info": { "name": "MiniCPM-V 2.0", "size": "-", "date": "2024-04-12", "type": "open_source", "link": "https://huggingface.co/openbmb/MiniCPM-V-2" }, "validation": { "overall": "37.1", "artDesign": "63.3", "business": "28.7", "science": "30.0", "healthMedicine": "30.0", "humanSocialSci": "56.7", "techEng": "27.1", "source": "author" } }, { "info": { "name": "LLaVA-1.5-13B", "size": "13B", "date": "2023-11-27", "type": "open_source", "link": "https://llava-vl.github.io/" }, "validation": { "overall": "36.4", "artDesign": "51.7", "business": "22.7", "science": "29.3", "healthMedicine": "38.7", "humanSocialSci": "53.3", "techEng": "31.4" }, "test": { "overall": "33.6", "artDesign": "49.7", "business": "28.2", "science": "25.9", "healthMedicine": "34.9", "humanSocialSci": "54.7", "techEng": "28.3" } }, { "info": { "name": "Emu2-Chat", "size": "-", "date": "2023-12-24", "type": "open_source", "link": "https://arxiv.org/abs/2312.13286" }, "validation": { "overall": "36.3", "artDesign": "55.0", "business": "30.0", "science": "28.7", "healthMedicine": "28.7", "humanSocialSci": "46.7", "techEng": "35.2", "source": "author" }, "test": { "overall": "34.1", "artDesign": "50.6", "business": "27.7", "science": "28.0", "healthMedicine": "32.4", "humanSocialSci": "50.3", "techEng": "31.3", "source": "author" } }, { "info": { "name": "Qwen-VL-7B-Chat", "size": "-", "date": "2023-11-27", "type": "open_source", "link": "https://arxiv.org/abs/2308.12966" }, "validation": { "overall": "35.9", "artDesign": "51.7", "business": "29.3", "science": "29.3", "healthMedicine": "33.3", "humanSocialSci": "45.0", "techEng": "32.9" }, "test": { "overall": "32.9", "artDesign": "47.6", "business": "29.8", "science": "25.6", "healthMedicine": "33.6", "humanSocialSci": "45.3", "techEng": "30.2" } }, { "info": { "name": "InstructBLIP-T5-XXL", "size": "-", "date": "2023-11-27", "type": "open_source", "link": "https://arxiv.org/abs/2305.06500" }, "validation": { "overall": "35.7", "artDesign": "44.2", "business": "24.0", "science": "30.7", "healthMedicine": "35.3", "humanSocialSci": "49.2", "techEng": "35.2" }, "test": { "overall": "33.8", "artDesign": "48.5", "business": "30.6", "science": "27.6", "healthMedicine": "33.6", "humanSocialSci": "49.8", "techEng": "29.4" } }, { "info": { "name": "BLIP-2 FLAN-T5-XXL", "size": "-", "date": "2023-11-27", "type": "open_source", "link": "https://arxiv.org/abs/2301.12597" }, "validation": { "overall": "35.4", "artDesign": "41.7", "business": "30.0", "science": "34.7", "healthMedicine": "32.0", "humanSocialSci": "50.8", "techEng": "30.0" }, "test": { "overall": "34.0", "artDesign": "49.2", "business": "28.6", "science": "27.3", "healthMedicine": "33.7", "humanSocialSci": "51.5", "techEng": "30.3" } }, { "info": { "name": "BLIP-2 FLAN-T5-XL", "size": "-", "date": "2023-11-27", "type": "open_source", "link": "https://arxiv.org/abs/2301.12597" }, "validation": { "overall": "34.4", "artDesign": "44.2", "business": "26.7", "science": "30.7", "healthMedicine": "35.3", "humanSocialSci": "50.0", "techEng": "27.6" }, "test": { "overall": "31.0", "artDesign": "43.0", "business": "25.6", "science": "25.1", "healthMedicine": "31.8", "humanSocialSci": "48.0", "techEng": "27.7" } }, { "info": { "name": "InstructBLIP-T5-XL", "size": "-", "date": "2023-11-27", "type": "open_source", "link": "https://arxiv.org/abs/2305.06500" }, "validation": { "overall": "32.9", "artDesign": "40.0", "business": "28.0", "science": "32.7", "healthMedicine": "28.7", "humanSocialSci": "47.5", "techEng": "27.1" }, "test": { "overall": "30.6", "artDesign": "43.3", "business": "25.2", "science": "25.2", "healthMedicine": "29.3", "humanSocialSci": "45.7", "techEng": "28.5" } }, { "info": { "name": "SPHINX", "size": "-", "date": "2023-12-26", "type": "open_source", "link": "https://arxiv.org/abs/2311.07575" }, "validation": { "overall": "32.9", "artDesign": "48.3", "business": "24.7", "science": "26.7", "healthMedicine": "30.7", "humanSocialSci": "50.0", "techEng": "26.2", "source": "author" }, "test": { "overall": "32.9", "artDesign": "50.8", "business": "27.2", "science": "25.3", "healthMedicine": "34.1", "humanSocialSci": "51.2", "techEng": "27.8", "source": "author" } }, { "info": { "name": "mPLUG-OWL2", "size": "-", "date": "2023-12-04", "type": "open_source", "link": "https://arxiv.org/abs/2311.04257" }, "validation": { "overall": "32.7", "artDesign": "45.8", "business": "24.7", "science": "22.7", "healthMedicine": "32.0", "humanSocialSci": "45.8", "techEng": "31.0", "source": "author" }, "test": { "overall": "32.1", "artDesign": "48.5", "business": "25.6", "science": "24.9", "healthMedicine": "32.8", "humanSocialSci": "46.7", "techEng": "29.5", "source": "author" } }, { "info": { "name": "Gemini Nano2", "size": "-", "date": "2023-12-06", "type": "proprietary", "link": "https://deepmind.google/technologies/gemini/#introduction" }, "validation": { "overall": "32.6", "source": "author" } }, { "info": { "name": "Otter", "size": "-", "date": "2023-11-27", "type": "open_source", "link": "https://arxiv.org/abs/2305.03726" }, "validation": { "overall": "32.2", "artDesign": "37.5", "business": "24.0", "science": "34.7", "healthMedicine": "30.7", "humanSocialSci": "41.7", "techEng": "29.0" }, "test": { "overall": "29.1", "artDesign": "37.4", "business": "24.0", "science": "24.1", "healthMedicine": "29.6", "humanSocialSci": "35.9", "techEng": "30.1" } }, { "info": { "name": "CogVLM", "size": "-", "date": "2023-11-27", "type": "open_source", "link": "https://arxiv.org/abs/2311.03079" }, "validation": { "overall": "32.1", "artDesign": "40.8", "business": "25.3", "science": "28.0", "healthMedicine": "32.0", "humanSocialSci": "45.0", "techEng": "27.6" }, "test": { "overall": "30.1", "artDesign": "37.9", "business": "25.6", "science": "25.1", "healthMedicine": "31.2", "humanSocialSci": "41.5", "techEng": "28.9" } }, { "info": { "name": "LLaMA-Adapter2-7B", "size": "7B", "date": "2023-11-27", "type": "open_source", "link": "https://arxiv.org/abs/2304.15010" }, "validation": { "overall": "29.8", "artDesign": "29.2", "business": "25.3", "science": "30.7", "healthMedicine": "30.7", "humanSocialSci": "33.3", "techEng": "30.0" }, "test": { "overall": "27.7", "artDesign": "35.2", "business": "25.4", "science": "25.6", "healthMedicine": "30.0", "humanSocialSci": "29.1", "techEng": "25.7" } }, { "info": { "name": "OpenFlamingo2-9B", "size": "9B", "date": "2023-11-27", "type": "open_source", "link": "https://arxiv.org/abs/2308.01390" }, "validation": { "overall": "28.7", "artDesign": "40.0", "business": "28.0", "science": "23.3", "healthMedicine": "27.3", "humanSocialSci": "30.8", "techEng": "26.2" }, "test": { "overall": "26.3", "artDesign": "31.6", "business": "23.5", "science": "26.3", "healthMedicine": "26.3", "humanSocialSci": "27.9", "techEng": "25.1" } }, { "info": { "name": "Adept Fuyu-8B", "size": "8B", "date": "2023-11-27", "type": "open_source", "link": "https://www.adept.ai/blog/fuyu-8b" }, "validation": { "overall": "27.9", "artDesign": "36.7", "business": "32.0", "science": "22.0", "healthMedicine": "28.0", "humanSocialSci": "32.5", "techEng": "21.4" }, "test": { "overall": "27.3", "artDesign": "29.9", "business": "27.0", "science": "25.6", "healthMedicine": "27.0", "humanSocialSci": "32.5", "techEng": "26.4" } }, { "info": { "name": "MiniGPT4-Vicuna-13B", "size": "13B", "date": "2023-11-27", "type": "open_source", "link": "https://minigpt-4.github.io/" }, "validation": { "overall": "26.8", "artDesign": "29.2", "business": "21.3", "science": "28.7", "healthMedicine": "30.7", "humanSocialSci": "29.2", "techEng": "23.8" }, "test": { "overall": "27.6", "artDesign": "30.2", "business": "27.0", "science": "26.2", "healthMedicine": "26.9", "humanSocialSci": "30.9", "techEng": "27.2" } }, { "info": { "name": "Kosmos2", "size": "-", "date": "2023-11-27", "type": "open_source", "link": "https://arxiv.org/abs/2306.14824" }, "validation": { "overall": "24.4", "artDesign": "25.0", "business": "18.0", "science": "19.3", "healthMedicine": "28.0", "humanSocialSci": "30.0", "techEng": "26.7" }, "test": { "overall": "26.6", "artDesign": "28.8", "business": "23.7", "science": "26.6", "healthMedicine": "27.2", "humanSocialSci": "26.3", "techEng": "26.8" } }, { "info": { "name": "InternVL2-Llama3-76B", "size": "76B", "date": "2024-07-04", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/InternVL2-Llama3-76B" }, "validation": { "overall": "58.3", "source": "author" }, "pro": { "overall": "40.0", "vision": "38.0", "original": "41.9" } },{ "info": { "name": "GPT-4o mini", "size": "-", "date": "2024-07-18", "type": "proprietary", "link": "https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/" }, "validation": { "overall": "59.4", "source": "author" }, "pro": { "overall": "37.6", "vision": "35.2", "original": "39.9" } }, { "info": { "name": "InternVL2-40B", "size": "40B", "date": "2024-07-04", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/InternVL2-40B" }, "validation": { "overall": "55.2", "source": "author" }, "pro": { "overall": "34.2", "vision": "32.1", "original": "36.3" } }, { "info": { "name": "LLaVA-OneVision-72B", "size": "72B", "date": "2024-08-06", "type": "open_source", "link": "https://huggingface.co/llava-hf/llava-onevision-qwen2-72b-ov-hf" }, "validation": { "overall": "56.8", "source": "author" }, "pro": { "overall": "31.0", "vision": "24.0", "original": "38.0" } }, { "info": { "name": "InternVL2-8B", "size": "8B", "date": "2024-07-04", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/InternVL2-8B" }, "validation": { "overall": "51.2", "source": "author" }, "pro": { "overall": "29.0", "vision": "25.4", "original": "32.5" } }, { "info": { "name": "MiniCPM-V 2.6", "size": "8B", "date": "2024-08-06", "type": "open_source", "link": "https://huggingface.co/openbmb/MiniCPM-V-2_6" }, "validation": { "overall": "49.8", "source": "author" }, "pro": { "overall": "27.2", "vision": "24.2", "original": "30.2" } }, { "info": { "name": "LlaVA-NEXT-72B", "size": "72B", "date": "2024-01-30", "type": "open_source", "link": "https://huggingface.co/llava-hf/llava-next-72b-hf" }, "validation": { "overall": "49.9", "source": "author" }, "pro": { "overall": "25.1", "vision": "19.2", "original": "31.0" } }, { "info": { "name": "LLaVA-OneVision-7B", "size": "7B", "date": "2024-08-06", "type": "open_source", "link": "https://huggingface.co/llava-hf/llava-onevision-qwen2-7b-ov-hf" }, "validation": { "overall": "48.8", "source": "author" }, "pro": { "overall": "24.1", "vision": "18.7", "original": "29.5" } }, { "info": { "name": "MiniCPM-Llama3-V 2.5", "size": "8.5B", "date": "2024-05-20", "type": "open_source", "link": "https://huggingface.co/openbmb/MiniCPM-Llama3-V-2_5" }, "validation": { "overall": "45.8", "source": "author" }, "pro": { "overall": "19.6", "vision": "14.5", "original": "24.6" } }, { "info": { "name": "Phi-3.5-Vision", "size": "-", "date": "2024-05-21", "type": "open_source", "link": "https://huggingface.co/microsoft/Phi-3-vision-128k-instruct" }, "validation": { "overall": "43.0", "source": "author" }, "pro": { "overall": "19.7", "vision": "13.1", "original": "26.3" } }, { "info": { "name": "LLaVA-NeXT-mistral-7B", "size": "7B", "date": "2024-01-30", "type": "open_source", "link": "https://huggingface.co/llava-hf/llava-v1.6-mistral-7b-hf" }, "validation": { "overall": "35.3", "source": "author" }, "pro": { "overall": "17.0", "vision": "14.6", "original": "19.4" } }, { "info": { "name": "LLaVA-NeXT-13B", "size": "13B", "date": "2024-01-30", "type": "open_source", "link": "https://huggingface.co/llava-hf/llava-v1.6-vicuna-13b-hf" }, "validation": { "overall": "36.2", "source": "author" }, "pro": { "overall": "17.2", "vision": "14.5", "original": "19.8" } }, { "info": { "name": "LLaVA-NeXT-Vicuna-7B", "size": "7B", "date": "2024-01-30", "type": "open_source", "link": "https://huggingface.co/llava-hf/llava-v1.6-vicuna-7b-hf" }, "validation": { "overall": "35.8", "source": "author" }, "pro": { "overall": "16.1", "vision": "13.4", "original": "18.8" } }, { "info": { "name": "Qwen2-VL-72B", "size": "72B", "date": "2024-08-29", "type": "proprietary", "link": "https://github.com/QwenLM/Qwen2-VL" }, "validation": { "overall": "64.5" }, "pro": { "overall": "46.2", "vision": "43.3", "original": "49.2" } }, { "info": { "name": "RBDash V1.2", "size": "72B", "date": "2024-08-21", "type": "open_source", "link": "https://github.com/RBDash-Team/RBDash" }, "validation": { "overall": "57.7", "artDesign": "68.3", "business": "52.0", "science": "53.3", "healthMedicine": "59.3", "humanSocialSci": "75.8", "techEng": "47.1", "source": "author" }, "test": { "overall": "51.2", "artDesign": "64.2", "business": "47.3", "science": "44.5", "healthMedicine": "57.0", "humanSocialSci": "74.9", "techEng": "41.8", "source": "author" } }, { "info": { "name": "360VL", "size": "70B", "date": "2024-05-21", "type": "open_source", "link": "https://github.com/360CVGroup/360VL" }, "validation": { "overall": "50.8", "source": "author" }, "test": { "overall": "44.3", "artDesign": "59.1", "business": "36.4", "science": "36.6", "healthMedicine": "49.2", "humanSocialSci": "70.4", "techEng": "36.9", "source": "author" } }, { "info": { "name": "Ovis-Clip-Llama3-8B", "size": "8B", "date": "2024-06-14", "type": "open_source", "link": "https://github.com/AIDC-AI/Ovis" }, "validation": { "overall": "48.6", "source": "author" }, "test": { "overall": "39.0", "artDesign": "54.3", "business": "33.3", "science": "31.7", "healthMedicine": "42.4", "humanSocialSci": "63.6", "techEng": "31.4", "source": "author" } }, { "info": { "name": "Ovis1.5-Gemma2-9B", "size": "9B", "date": "2024-06-14", "type": "open_source", "link": "https://github.com/AIDC-AI/Ovis" }, "validation": { "overall": "49.8", "source": "author" } }, { "info": { "name": "Gemma 3-4B", "size": "4B", "date": "2025-03-12", "type": "open_source", "link": "https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf" }, "validation": { "overall": "48.8", "source": "author" } }, { "info": { "name": "Gemma 3-12B", "size": "12B", "date": "2025-03-12", "type": "open_source", "link": "https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf" }, "validation": { "overall": "59.6", "source": "author" } }, { "info": { "name": "Gemma 3-27B", "size": "27B", "date": "2025-03-12", "type": "open_source", "link": "https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf" }, "validation": { "overall": "64.9", "source": "author" } }, { "info": { "name": "Kimi-VL-A3B", "size": "16B", "date": "2025-04-10", "type": "open_source", "link": "https://github.com/MoonshotAI/Kimi-VL" }, "validation": { "overall": "57.0", "source": "author" } }, { "info": { "name": "NVLM-D 1.0 78B", "size": "78B", "date": "2024-09-17", "type": "open_source", "link": "https://nvlm-project.github.io/" }, "validation": { "overall": "59.7", "source": "author" }, "test": { "overall": "54.6", "artDesign": "66.0", "business": "49.4", "science": "47.3", "healthMedicine": "60.5", "humanSocialSci": "77.8", "techEng": "47.0", "source": "author" } }, { "info": { "name": "NVLM-H 1.0 78B", "size": "78B", "date": "2024-09-17", "type": "open_source", "link": "https://nvlm-project.github.io/" }, "validation": { "overall": "60.2", "source": "author" }, "test": { "overall": "53.0", "artDesign": "64.8", "business": "48.9", "science": "46.4", "healthMedicine": "58.7", "humanSocialSci": "73.9", "techEng": "45.2", "source": "author" } }, { "info": { "name": "Llama 3.2 11B", "size": "11B", "date": "2024-09-25", "type": "open_source", "link": "https://ai.meta.com/blog/llama-3-2-connect-2024-vision-edge-mobile-devices/" }, "validation": { "overall": "50.7", "source": "author" }, "pro": { "overall": "28.4", "vision": "23.7", "original": "33.0", "source": "author" } }, { "info": { "name": "Llama 3.2 90B", "size": "90B", "date": "2024-09-25", "type": "open_source", "link": "https://ai.meta.com/blog/llama-3-2-connect-2024-vision-edge-mobile-devices/" }, "validation": { "overall": "60.3", "source": "author" }, "pro": { "overall": "39.5", "vision": "33.8", "original": "45.2", "source": "author" } }, { "info": { "name": "Llama 4 Behemoth", "size": "288B", "date": "2025-04-05", "type": "proprietary", "link": "https://ai.meta.com/blog/llama-4-multimodal-intelligence/" }, "validation": { "overall": "76.1", "source": "author" } }, { "info": { "name": "Llama 4 Maverick", "size": "17B", "date": "2025-04-05", "type": "open_source", "link": "https://ai.meta.com/blog/llama-4-multimodal-intelligence/" }, "validation": { "overall": "73.4", "source": "author" } }, { "info": { "name": "Llama 4 Scout", "size": "17B", "date": "2025-04-05", "type": "open_source", "link": "https://ai.meta.com/blog/llama-4-multimodal-intelligence/" }, "validation": { "overall": "69.4", "source": "author" } }, { "info": { "name": "TeleMM", "size": "-", "date": "2024-11-18", "type": "proprietary", "link": "https://cloud.siliconflow.cn/playground/chat/17885302607" }, "validation": { "overall": "61.4", "artDesign": "68.3", "business": "62.7", "science": "52.0", "healthMedicine": "65.3", "humanSocialSci": "77.6", "techEng": "51.4", "source": "author" }, "test": { "overall": "58.2", "artDesign": "69.4", "business": "61.8", "science": "51.6", "healthMedicine": "65.9", "humanSocialSci": "78.1", "techEng": "45.9", "source": "author" } }, { "info": { "name": "MAmmoTH-VL-8B", "size": "8B", "date": "2024-12-08", "type": "open_source", "link": "https://mammoth-vl.github.io/" }, "validation": { "overall": "50.8", "source": "author" }, "pro": { "overall": "25.3", "source": "author" } }, { "info": { "name": "InternVL2.5-78B", "size": "78B", "date": "2024-12-05", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/InternVL2_5-78B" }, "validation": { "overall": "70.1", "source": "author" }, "test": { "overall": "61.8", "source": "author" }, "pro": { "overall": "48.6", "vision": "45.9", "original": "51.4", "source": "author" } }, { "info": { "name": "InternVL2.5-38B", "size": "38B", "date": "2024-12-05", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/InternVL2_5-38B" }, "validation": { "overall": "63.9", "source": "author" }, "test": { "overall": "57.6", "source": "author" }, "pro": { "overall": "46.0", "vision": "44.0", "original": "48.0", "source": "author" } }, { "info": { "name": "InternVL2.5-26B", "size": "26B", "date": "2024-12-05", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/InternVL2_5-26B" }, "validation": { "overall": "60.0", "source": "author" }, "test": { "overall": "51.8", "source": "author" }, "pro": { "overall": "37.1", "vision": "32.6", "original": "41.6", "source": "author" } }, { "info": { "name": "InternVL2.5-8B", "size": "8B", "date": "2024-12-05", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/InternVL2_5-8B" }, "validation": { "overall": "56.0", "source": "author" }, "test": { "overall": "48.9", "source": "author" }, "pro": { "overall": "34.3", "vision": "30.4", "original": "38.2", "source": "author" } }, { "info": { "name": "InternVL2.5-2B", "size": "2B", "date": "2024-12-05", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/InternVL2_5-2B" }, "validation": { "overall": "43.6", "source": "author" }, "test": { "overall": "38.2", "source": "author" }, "pro": { "overall": "23.7", "vision": "20.1", "original": "27.3", "source": "author" } }, { "info": { "name": "InternVL2.5-1B", "size": "1B", "date": "2024-12-05", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/InternVL2_5-1B" }, "validation": { "overall": "40.9", "source": "author" }, "test": { "overall": "35.8", "source": "author" }, "pro": { "overall": "19.4", "vision": "15.5", "original": "23.3", "source": "author" } }, { "info": { "name": "QVQ-72B-Preview", "size": "72B", "date": "2024-12-25", "type": "open_source", "link": "https://qwenlm.github.io/blog/qvq-72b-preview/" }, "validation": { "overall": "70.3", "source": "author" } }, { "info": { "name": "Qwen2.5-VL 3B", "size": "3B", "date": "2025-02-19", "type": "open_source", "link": "https://arxiv.org/pdf/2502.13923" }, "validation": { "overall": "53.1", "source": "author" }, "pro": { "overall": "31.6", "source": "author" } }, { "info": { "name": "Qwen2.5-VL 7B", "size": "7B", "date": "2025-02-19", "type": "open_source", "link": "https://arxiv.org/pdf/2502.13923" }, "validation": { "overall": "58.6", "source": "author" }, "pro": { "overall": "38.3", "source": "author" } }, { "info": { "name": "Qwen2.5-VL 72B", "size": "72B", "date": "2025-02-19", "type": "open_source", "link": "https://arxiv.org/pdf/2502.13923" }, "validation": { "overall": "64.5", "source": "author" }, "pro": { "overall": "46.2", "source": "author" } }, { "info": { "name": "HoVLE", "size": "2.6B", "date": "2024-12-18", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/HoVLE" }, "validation": { "overall": "33.7", "source": "author" } }, { "info": { "name": "HoVLE (HD)", "size": "2.6B", "date": "2024-12-18", "type": "open_source", "link": "https://huggingface.co/OpenGVLab/HoVLE-HD" }, "validation": { "overall": "33.2", "source": "author" } }, { "info": { "name": "LLaVA-UHD v2", "size": "-", "date": "2024-12-21", "type": "open_source", "link": "https://arxiv.org/pdf/2412.13871" }, "validation": { "overall": "38.2", "source": "author" } }, { "info": { "name": "MetaMorph", "size": "-", "date": "2024-12-19", "type": "proprietary", "link": "https://tsb0601.github.io/metamorph/" }, "validation": { "overall": "41.8", "source": "author" } }, { "info": { "name": "ILLUME", "size": "-", "date": "2024-12-09", "type": "proprietary", "link": "https://arxiv.org/pdf/2412.06673" }, "validation": { "overall": "38.2", "source": "author" } }, { "info": { "name": "Emu3", "size": "8B", "date": "2024-09-27", "type": "open_source", "link": "https://emu.baai.ac.cn/about" }, "validation": { "overall": "31.6", "source": "author" } }, { "info": { "name": "POINTS1.5", "size": "7B", "date": "2024-12-11", "type": "open_source", "link": "https://github.com/WePOINTS/WePOINTS?tab=readme-ov-file" }, "validation": { "overall": "53.8", "source": "author" } }, { "info": { "name": "HyViLM", "size": "8B", "date": "2024-12-11", "type": "proprietary", "link": "https://arxiv.org/abs/2412.08378" }, "validation": { "overall": "41.8", "source": "author" } }, { "info": { "name": "Olympus", "size": "2.7B", "date": "2024-12-12", "type": "open_source", "link": "https://yuanze-lin.me/Olympus_page/" }, "validation": { "overall": "32.8", "source": "author" } }, { "info": { "name": "SynerGen-VL", "size": "2.4B", "date": "2024-12-12", "type": "proprietary", "link": "https://arxiv.org/abs/2412.09604" }, "validation": { "overall": "34.2", "source": "author" } }, { "info": { "name": "Show-o", "size": "1.3B", "date": "2024-08-22", "type": "open_source", "link": "https://showlab.github.io/Show-o/" }, "validation": { "overall": "27.4", "source": "author" } }, { "info": { "name": "EVLM-KTO", "size": "7B", "date": "2024-12-13", "type": "proprietary", "link": "https://arxiv.org/abs/2412.10566" }, "validation": { "overall": "53.0", "source": "author" }, "pro": { "overall": "43.8", "source": "author" } }, { "info": { "name": "DreamPRM-1.5 (GPT-5-mini w/ thinking)", "size": "-", "date": "2025-09-05", "type": "proprietary", "link": "https://github.com/coder-qicao/DreamPRM-1.5" }, "validation": { "overall": "84.6", "source": "author" } }, { "info": { "name": "MiMo-VL 7B-SFT", "size": "7B", "date": "2025-06-04", "type": "open_source", "link": "https://github.com/XiaomiMiMo/MiMo-VL" }, "validation": { "overall": "64.6", "source": "author" }, "pro": { "overall": "42.3", "vision": "39.4", "original": "45.2", "source": "author" } }, { "info": { "name": "MiMo-VL 7B-RL", "size": "7B", "date": "2025-06-04", "type": "open_source", "link": "https://github.com/XiaomiMiMo/MiMo-VL" }, "validation": { "overall": "66.7", "source": "author" }, "pro": { "overall": "43.3", "vision": "40.3", "original": "46.2", "source": "author" } }, { "info": { "name": "MiMo-VL 7B-RL-2508 w/ Thinking", "size": "7B", "date": "2025-08-08", "type": "open_source", "link": "https://github.com/XiaomiMiMo/MiMo-VL" }, "validation": { "overall": "70.6", "source": "author" } }, { "info": { "name": "MiMo-VL 7B-RL-2508 w/o Thinking", "size": "7B", "date": "2025-08-08", "type": "open_source", "link": "https://github.com/XiaomiMiMo/MiMo-VL" }, "validation": { "overall": "63.1", "source": "author" } }, { "info": { "name": "GLM-4.5V w/ Thinking", "size": "12B", "date": "2025-08-11", "type": "open_source", "link": "https://docs.z.ai/guides/vlm/glm-4.5v" }, "validation": { "overall": "75.4", "source": "author" }, "pro": { "overall": "65.2", "source": "author" } }, { "info": { "name": "GLM-4.1V w/ Thinking", "size": "9B", "date": "2025-07-01", "type": "open_source", "link": "https://arxiv.org/abs/2507.01006" }, "validation": { "overall": "68.0", "source": "author" }, "pro": { "overall": "57.1", "source": "author" } }, { "info": { "name": "DeepSeek-VL2", "size": "4.5B", "date": "2024-12-13", "type": "open_source", "link": "https://github.com/deepseek-ai/DeepSeek-VL2" }, "validation": { "overall": "51.1", "vision": "42.0", "standard": "60.2", "source": "author" } }, { "info": { "name": "DeepSeek-VL2-Small", "size": "2.8B", "date": "2024-12-13", "type": "open_source", "link": "https://github.com/deepseek-ai/DeepSeek-VL2" }, "validation": { "overall": "48.0", "source": "author" } }, { "info": { "name": "DeepSeek-VL2-Tiny", "size": "1.0B", "date": "2024-12-13", "type": "open_source", "link": "https://github.com/deepseek-ai/DeepSeek-VL2" }, "validation": { "overall": "40.7", "source": "author" } }, { "info": { "name": "Seed 1.6-Thinking", "size": "20B", "date": "2025-06-25", "type": "proprietary", "link": "https://seed.bytedance.com/en/seed1_6" }, "validation": { "overall": "74.8", "source": "author" }, "pro": { "overall": "66.4", "source": "author" } }, { "info": { "name": "Seed 1.5-VL", "size": "20B", "date": "2025-05-13", "type": "proprietary", "link": "https://seed.bytedance.com/en/tech/seed1_5_vl" }, "validation": { "overall": "73.6", "source": "author" }, "pro": { "overall": "59.9", "source": "author" } }, { "info": { "name": "Seed 1.5-VL Thinking", "size": "20B", "date": "2025-05-13", "type": "proprietary", "link": "https://seed.bytedance.com/en/tech/seed1_5_vl" }, "validation": { "overall": "77.9", "source": "author" }, "pro": { "overall": "67.6", "source": "author" } }, { "info": { "name": "Aria", "size": "3.9B", "date": "2024-10-10", "type": "open_source", "link": "https://rhymes.ai/blog-details/aria-first-open-multimodal-native-moe-model" }, "validation": { "overall": "54.9", "source": "author" } }, { "info": { "name": "Spark4.0-Turbo", "size": "-", "date": "2025-01-22", "type": "proprietary", "link": "https://xinghuo.xfyun.cn/spark" }, "validation": { "overall": "70.4", "artDesign": "71.7", "business": "76.0", "science": "62.7", "healthMedicine": "78.0", "humanSocialSci": "85.0", "techEng": "57.6", "source": "author" }, "test": { "overall": "60.3", "artDesign": "67.4", "business": "66.8", "science": "53.0", "healthMedicine": "68.4", "humanSocialSci": "80.3", "techEng": "48.6", "source": "author" } }, { "info": { "name": "InternVL3-1B", "size": "1B", "date": "2025-04-11", "type": "open_source", "link": "https://internvl.github.io/blog/2025-04-11-InternVL-3.0/" }, "validation": { "overall": "43.4", "source": "author" } }, { "info": { "name": "InternVL3-2B", "size": "2B", "date": "2025-04-11", "type": "open_source", "link": "https://internvl.github.io/blog/2025-04-11-InternVL-3.0/" }, "validation": { "overall": "48.6", "source": "author" } }, { "info": { "name": "InternVL3-8B", "size": "8B", "date": "2025-04-11", "type": "open_source", "link": "https://internvl.github.io/blog/2025-04-11-InternVL-3.0/" }, "validation": { "overall": "62.7", "source": "author" } }, { "info": { "name": "InternVL3-9B", "size": "9B", "date": "2025-04-11", "type": "open_source", "link": "https://internvl.github.io/blog/2025-04-11-InternVL-3.0/" }, "validation": { "overall": "57.7", "source": "author" } }, { "info": { "name": "InternVL3-14B", "size": "14B", "date": "2025-04-11", "type": "open_source", "link": "https://internvl.github.io/blog/2025-04-11-InternVL-3.0/" }, "validation": { "overall": "67.1", "source": "author" } }, { "info": { "name": "InternVL3-38B", "size": "38B", "date": "2025-04-11", "type": "open_source", "link": "https://internvl.github.io/blog/2025-04-11-InternVL-3.0/" }, "validation": { "overall": "70.1", "source": "author" } }, { "info": { "name": "InternVL3-78B", "size": "78B", "date": "2025-04-11", "type": "open_source", "link": "https://internvl.github.io/blog/2025-04-11-InternVL-3.0/" }, "validation": { "overall": "72.2", "source": "author" } }, { "info": { "name": "Ovis2-1B", "size": "1B", "date": "2025-03-25", "type": "open_source", "link": "https://github.com/AIDC-AI/Ovis" }, "validation": { "overall": "36.0", "source": "author" } }, { "info": { "name": "Ovis2-2B", "size": "2B", "date": "2025-03-25", "type": "open_source", "link": "https://github.com/AIDC-AI/Ovis" }, "validation": { "overall": "45.3", "source": "author" } }, { "info": { "name": "Ovis2-4B", "size": "4B", "date": "2025-03-25", "type": "open_source", "link": "https://github.com/AIDC-AI/Ovis" }, "validation": { "overall": "48.0", "source": "author" } }, { "info": { "name": "Ovis2-8B", "size": "8B", "date": "2025-03-25", "type": "open_source", "link": "https://github.com/AIDC-AI/Ovis" }, "validation": { "overall": "59.0", "source": "author" } }, { "info": { "name": "Ovis2-16B", "size": "16B", "date": "2025-03-25", "type": "open_source", "link": "https://github.com/AIDC-AI/Ovis" }, "validation": { "overall": "59.6", "source": "author" } }, { "info": { "name": "Ovis2-34B", "size": "34B", "date": "2025-03-25", "type": "open_source", "link": "https://github.com/AIDC-AI/Ovis" }, "validation": { "overall": "65.6", "source": "author" } }, { "info": { "name": "Cambrian-1-8B", "size": "8B", "date": "2024-06-27", "type": "open_source", "link": "https://cambrian-mllm.github.io/" }, "validation": { "overall": "42.7", "source": "author" } }, { "info": { "name": "Cambrian-1-13B", "size": "13B", "date": "2024-06-27", "type": "open_source", "link": "https://cambrian-mllm.github.io/" }, "validation": { "overall": "40.0", "source": "author" } }, { "info": { "name": "Cambrian-1-34B", "size": "34B", "date": "2024-06-27", "type": "open_source", "link": "https://cambrian-mllm.github.io/" }, "validation": { "overall": "49.7", "source": "author" } }, { "info": { "name": "Aquila-VL-2B", "size": "2B", "date": "2024-10-25", "type": "open_source", "link": "https://huggingface.co/BAAI/Aquila-VL-2B-llava-qwen" }, "validation": { "overall": "47.4", "source": "author" } }, { "info": { "name": "Skywork-R1V3-38B", "size": "38B", "date": "2025-07-05", "type": "open_source", "link": "https://huggingface.co/Skywork/Skywork-R1V3-38B" }, "validation": { "overall": "76.0", "source": "author" }, "pro": { "overall": "55.4", "source": "author" } }, { "info": { "name": "Skywork-R1V2-38B", "size": "38B", "date": "2025-04-28", "type": "open_source", "link": "https://huggingface.co/Skywork/Skywork-R1V2-38B-AWQ" }, "validation": { "overall": "73.6", "source": "author" } }, { "info": { "name": "Skywork-R1V-38B", "size": "38B", "date": "2025-04-28", "type": "open_source", "link": "https://huggingface.co/Skywork/Skywork-R1V-38B-AWQ" }, "validation": { "overall": "69.0", "source": "author" } }, { "info": { "name": "Eagle2.5-8B", "size": "8B", "date": "2025-04-21", "type": "open_source", "link": "https://nvlabs.github.io/EAGLE/" }, "validation": { "overall": "55.8", "source": "author" } }, { "info": { "name": "BLIP3-o4B", "size": "4B", "date": "2025-05-16", "type": "open_source", "link": "https://github.com/JiuhaiChen/BLIP3o" }, "validation": { "overall": "46.6", "source": "author" } }, { "info": { "name": "BLIP3-o8B", "size": "8B", "date": "2025-05-16", "type": "open_source", "link": "https://github.com/JiuhaiChen/BLIP3o" }, "validation": { "overall": "50.6", "source": "author" } }, { "info": { "name": "Flash-VL-2B-Dynamic", "size": "2B", "date": "2025-05-16", "type": "open_source", "link": "https://huggingface.co/FlashVL/FlashVL-2B-Dynamic" }, "validation": { "overall": "42.9", "source": "author" } }, { "info": { "name": "Aya-Vision-8B", "size": "8B", "date": "2025-05-13", "type": "open_source", "link": "https://huggingface.co/CohereLabs/aya-vision-8b" }, "validation": { "overall": "39.9", "source": "author" } }, { "info": { "name": "Aya-Vision-32B", "size": "32B", "date": "2025-05-13", "type": "open_source", "link": "https://huggingface.co/CohereLabs/aya-vision-32b" }, "validation": { "overall": "45.11", "source": "author" } }, { "info": { "name": "Mogao-7B", "size": "8B", "date": "2025-05-08", "type": "open_source", "link": "https://arxiv.org/abs/2505.05472" }, "validation": { "overall": "44.2", "source": "author" } }, { "info": { "name": "Nova Premier", "size": "-", "date": "2025-04-30", "type": "proprietary", "link": "https://www.amazon.science/publications/amazon-nova-premier-technical-report-and-model-card" }, "validation": { "overall": "68.0", "source": "author" } }, { "info": { "name": "Nova Pro", "size": "-", "date": "2025-04-30", "type": "proprietary", "link": "https://www.amazon.science/publications/amazon-nova-premier-technical-report-and-model-card" }, "validation": { "overall": "62.0", "source": "author" } }, { "info": { "name": "SpaRE-2B", "size": "2B", "date": "2025-04-29", "type": "open_source", "link": "https://arxiv.org/abs/2504.20648" }, "validation": { "overall": "40.0", "source": "author" } }, { "info": { "name": "SpaRE-7B", "size": "7B", "date": "2025-04-29", "type": "open_source", "link": "https://arxiv.org/abs/2504.20648" }, "validation": { "overall": "51.0", "source": "author" } }, { "info": { "name": "Frequent Choice", "size": "-", "date": "2023-11-27", "type": "random_frequent" }, "validation": { "overall": "26.8", "artDesign": "23.3", "business": "29.3", "science": "27.3", "healthMedicine": "30.0", "humanSocialSci": "25.8", "techEng": "24.8" }, "test": { "overall": "23.9", "artDesign": "24.1", "business": "24.9", "science": "21.6", "healthMedicine": "25.3", "humanSocialSci": "22.8", "techEng": "24.8" }, "pro": { "overall": "12.1", "vision": "12.1", "original": "12.1" } }, { "info": { "name": "Random Choice", "size": "-", "date": "2023-11-27", "type": "random_frequent" }, "validation": { "overall": "22.1", "artDesign": "29.2", "business": "24.7", "science": "18.0", "healthMedicine": "20.7", "humanSocialSci": "20.0", "techEng": "21.4" }, "test": { "overall": "25.8", "artDesign": "26.7", "business": "28.4", "science": "24.0", "healthMedicine": "24.4", "humanSocialSci": "25.2", "techEng": "26.5" }, "pro": { "overall": "12.6", "vision": "12.4", "original": "12.8" } } ] }