New: connect Claude & other AIs to GenAIList over MCP โ research the catalog and contribute to the shared knowledge base. Learn how โ
Multimodal AI models
Explore multimodal AI models that understand images, text, audio and video together. Compare vision-language models (VLMs) by multimodal benchmark scores and capabilities.
Llama 4 Maverick
๐บ๐ธ Meta AI
Llama 4 Scout
๐บ๐ธ Meta AI
GPT-4o (Mar 2025)
๐บ๐ธ OpenAI
Qwen2.5-Omni 3B
๐จ๐ณ Qwen
Qwen2.5-Omni 7B
๐จ๐ณ Qwen
ERNIE-4.5-VL-424B-A47B (ๆๅฟๅคงๆจกๅ4.5)
๐จ๐ณ Baidu
Reka Flash 3
๐บ๐ธ Reka AI
Mistral OCR
๐ซ๐ท Mistral AI
Aya Vision 32B
๐จ๐ฆ Cohere
Phi-4-Multimodal
๐บ๐ธ Microsoft
TasiChat (TasiChatๅคงๆจกๅ)
๐จ๐ณ "Chengdu Tasi Technology Co.
YAYI-Ultra
๐จ๐ณ Yayi (Wenge)
GPT-4o (Jan 2025)
๐บ๐ธ OpenAI
Baichuan-Omni-1.5
๐จ๐ณ Baichuan
Kimi k1.5
๐จ๐ณ Moonshot
LinGan VL (ไธดๆVL)
๐จ๐ณ Beijing 58 Information Technology
Apollo-7B
๐บ๐ธ Meta AI
InternVL2_5-26B
๐จ๐ณ Shanghai AI Lab
InternVL2_5-38B
๐จ๐ณ Shanghai AI Lab
InternVL2_5-78B
๐จ๐ณ Shanghai AI Lab
Genie 2
๐บ๐ธ Google DeepMind
Amazon Nova Lite
๐บ๐ธ Amazon
Amazon Nova Pro
๐บ๐ธ Amazon
Intelligent Go-Explore (IGE)
๐จ๐ฆ University of British Columbia (UBC)
Ovis1.6-Gemma2-27B
๐จ๐ณ Alibaba
Fugatto 1
๐บ๐ธ NVIDIA
Gauss2
๐ฐ๐ท Samsung
GPT-4o (Nov 2024)
๐บ๐ธ OpenAI
Pixtral Large
๐ซ๐ท Mistral AI
Aha (ๅคงๆจกๅ)
๐จ๐ณ "Shanghai Xingzhi Technology Co.
LongVU
๐บ๐ธ Meta AI
Taiyi (ๆท่งๅคชไน)
๐จ๐ณ Megvii Inc
Zhiyun Culture LLM (ๆบไบๆๅๅคงๆจกๅ)
๐จ๐ณ "Xinhua Zhiyun Technology Co.
Aria
๐บ๐ธ Rhymes AI
Gemini 1.5 Flash (Sep 2024)
๐บ๐ธ Google DeepMind
Llama 3.2 11B
๐บ๐ธ Meta AI
Llama 3.2 90B
๐บ๐ธ Meta AI
Oryx 34B
๐จ๐ณ Tsinghua University
Oryx 7B
๐จ๐ณ Tsinghua University
Ovis1.6-Gemma2-9B
๐จ๐ณ Alibaba
Galaxy (ๆๆฑGalaxyๅคงๆจกๅ)
๐จ๐ณ Dahan Software / Hanweb
LLaVA-MORE 8B
๐ฎ๐น AImageLab
GameNGen
๐บ๐ธ Google Research
Kosmos-2.5
๐บ๐ธ Microsoft
LongVILA-7B
๐บ๐ธ NVIDIA
xGen-MM (BLIP-3)
๐บ๐ธ Salesforce Research
Zhiye LLM (ๆตชๆฝฎ็ฅไธๅคงๆจกๅ)
๐จ๐ณ Inspur
GPT-4o (Aug 2024)
๐บ๐ธ OpenAI
LLaVA-OV-72B
๐จ๐ณ ByteDance
LLaVA-OV-7B
๐จ๐ณ ByteDance
Xingchen Multimodal Model (ไธญ็ตไฟกไบบๅทฅๆบ่ฝ็งๆ๏ผๅไบฌ๏ผๆ้ๅ ฌๅธ)
๐จ๐ณ China Telecom
LLaVA-NeXT-32B-Qwen
LMMs-Lab
LLaVA-NeXT-32B-Qwen
๐จ๐ณ Qwen
InternVL2-Llama3-76B
๐จ๐ณ Shanghai AI Lab
Qwen2-Audio
๐จ๐ณ Qwen
SenseChat 5.5
๐จ๐ณ SenseTime
InternVL2 26B
๐จ๐ณ Shanghai AI Lab
InternVL2-40B
๐จ๐ณ Shanghai AI Lab
Ernie 4.0 Turbo
๐จ๐ณ Baidu
Cambrian-1-13B
๐บ๐ธ New York University (NYU)
About Multimodal AI models
Multimodal AI models โ often called vision-language models (VLMs) โ understand more than text: they reason over images, documents, charts, audio and sometimes video alongside language. The best multimodal model depends on your inputs: document understanding rewards strong OCR and layout reasoning, while visual question answering rewards fine-grained image understanding. When comparing VLMs, look at multimodal benchmark scores, supported input types, resolution limits and context window, then weigh quality against cost and latency. Browse the list below to compare multimodal models by provider and availability, check rankings on our benchmarks page, and use compare to put two vision-language models side by side.
Frequently asked questions
How do I choose the right model?
Weigh raw capability against practical constraints like context window, latency, licensing and price. Use the benchmarks page to compare rankings and the compare tool to evaluate two candidates side by side.
Where can I see benchmark scores?
Visit the benchmarks page to compare these models on standardised tests, then use the compare tool for a detailed side-by-side of any two models.