// MODEL CATEGORY

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.

271 models
MA

Llama 4 Maverick

๐Ÿ‡บ๐Ÿ‡ธ Meta AI

Apr 2025 400B
Multimodal Open (Restricted)
MA

Llama 4 Scout

๐Ÿ‡บ๐Ÿ‡ธ Meta AI

Apr 2025 109B
Multimodal Open (Restricted)
OpenAI

GPT-4o (Mar 2025)

๐Ÿ‡บ๐Ÿ‡ธ OpenAI

Mar 2025
Multimodal Proprietary
Qwen

Qwen2.5-Omni 3B

๐Ÿ‡จ๐Ÿ‡ณ Qwen

Mar 2025 3B
Multimodal Open Weights
Qwen

Qwen2.5-Omni 7B

๐Ÿ‡จ๐Ÿ‡ณ Qwen

Mar 2025 7B
Multimodal Open (Restricted)
BA

ERNIE-4.5-VL-424B-A47B (ๆ–‡ๅฟƒๅคงๆจกๅž‹4.5)

๐Ÿ‡จ๐Ÿ‡ณ Baidu

Mar 2025 424B
Multimodal Open (Restricted)
RA

Reka Flash 3

๐Ÿ‡บ๐Ÿ‡ธ Reka AI

Mar 2025 21B
Multimodal Open (Restricted)
Mistral AI

Mistral OCR

๐Ÿ‡ซ๐Ÿ‡ท Mistral AI

Mar 2025
Multimodal Proprietary
Cohere

Aya Vision 32B

๐Ÿ‡จ๐Ÿ‡ฆ Cohere

Mar 2025 33.1B
Multimodal Open Weights
Microsoft

Phi-4-Multimodal

๐Ÿ‡บ๐Ÿ‡ธ Microsoft

Mar 2025 5.6B
Multimodal Open (Restricted)
"T

TasiChat (TasiChatๅคงๆจกๅž‹)

๐Ÿ‡จ๐Ÿ‡ณ "Chengdu Tasi Technology Co.

Feb 2025
Multimodal Proprietary
Y(

YAYI-Ultra

๐Ÿ‡จ๐Ÿ‡ณ Yayi (Wenge)

Feb 2025
Multimodal Proprietary
OpenAI

GPT-4o (Jan 2025)

๐Ÿ‡บ๐Ÿ‡ธ OpenAI

Jan 2025
Multimodal Proprietary
Baichuan

Baichuan-Omni-1.5

๐Ÿ‡จ๐Ÿ‡ณ Baichuan

Jan 2025 11B
Multimodal Open (Restricted)
MO

Kimi k1.5

๐Ÿ‡จ๐Ÿ‡ณ Moonshot

Jan 2025
Multimodal Proprietary
B5

LinGan VL (ไธดๆ„ŸVL)

๐Ÿ‡จ๐Ÿ‡ณ Beijing 58 Information Technology

Dec 2024
Multimodal Proprietary
MA

Apollo-7B

๐Ÿ‡บ๐Ÿ‡ธ Meta AI

Dec 2024 7B
Multimodal Proprietary
SA

InternVL2_5-26B

๐Ÿ‡จ๐Ÿ‡ณ Shanghai AI Lab

Dec 2024 25.5B
Multimodal Open (Restricted)
SA

InternVL2_5-38B

๐Ÿ‡จ๐Ÿ‡ณ Shanghai AI Lab

Dec 2024 38.4B
Multimodal Open (Restricted)
SA

InternVL2_5-78B

๐Ÿ‡จ๐Ÿ‡ณ Shanghai AI Lab

Dec 2024 78.4B
Multimodal Open (Restricted)
Google DeepMind

Genie 2

๐Ÿ‡บ๐Ÿ‡ธ Google DeepMind

Dec 2024
Multimodal Proprietary
AM

Amazon Nova Lite

๐Ÿ‡บ๐Ÿ‡ธ Amazon

Dec 2024
Multimodal Proprietary
AM

Amazon Nova Pro

๐Ÿ‡บ๐Ÿ‡ธ Amazon

Dec 2024
Multimodal Proprietary
UO

Intelligent Go-Explore (IGE)

๐Ÿ‡จ๐Ÿ‡ฆ University of British Columbia (UBC)

Dec 2024
Multimodal Open (Restricted)
AL

Ovis1.6-Gemma2-27B

๐Ÿ‡จ๐Ÿ‡ณ Alibaba

Nov 2024 28.9B
Multimodal Open (Restricted)
NVIDIA

Fugatto 1

๐Ÿ‡บ๐Ÿ‡ธ NVIDIA

Nov 2024 2.5B
Multimodal Proprietary
SA

Gauss2

๐Ÿ‡ฐ๐Ÿ‡ท Samsung

Nov 2024
Multimodal Proprietary
OpenAI

GPT-4o (Nov 2024)

๐Ÿ‡บ๐Ÿ‡ธ OpenAI

Nov 2024
Multimodal Proprietary
Mistral AI

Pixtral Large

๐Ÿ‡ซ๐Ÿ‡ท Mistral AI

Nov 2024 124B
Multimodal Open Weights
"X

Aha (ๅคงๆจกๅž‹)

๐Ÿ‡จ๐Ÿ‡ณ "Shanghai Xingzhi Technology Co.

Oct 2024
Multimodal Proprietary
MA

LongVU

๐Ÿ‡บ๐Ÿ‡ธ Meta AI

Oct 2024 7B
Multimodal Open (Restricted)
MI

Taiyi (ๆ—ท่ง†ๅคชไน™)

๐Ÿ‡จ๐Ÿ‡ณ Megvii Inc

Oct 2024
Multimodal Proprietary
"Z

Zhiyun Culture LLM (ๆ™บไบ‘ๆ–‡ๅŒ–ๅคงๆจกๅž‹)

๐Ÿ‡จ๐Ÿ‡ณ "Xinhua Zhiyun Technology Co.

Oct 2024
Multimodal Proprietary
RA

Aria

๐Ÿ‡บ๐Ÿ‡ธ Rhymes AI

Oct 2024 24.9B
Multimodal Open (Restricted)
Google DeepMind

Gemini 1.5 Flash (Sep 2024)

๐Ÿ‡บ๐Ÿ‡ธ Google DeepMind

Sep 2024
Multimodal Proprietary
MA

Llama 3.2 11B

๐Ÿ‡บ๐Ÿ‡ธ Meta AI

Sep 2024 10.6B
Multimodal Open (Restricted)
MA

Llama 3.2 90B

๐Ÿ‡บ๐Ÿ‡ธ Meta AI

Sep 2024 88.6B
Multimodal Open (Restricted)
TU

Oryx 34B

๐Ÿ‡จ๐Ÿ‡ณ Tsinghua University

Sep 2024 34B
Multimodal Open (Restricted)
TU

Oryx 7B

๐Ÿ‡จ๐Ÿ‡ณ Tsinghua University

Sep 2024 7B
Multimodal Open (Restricted)
AL

Ovis1.6-Gemma2-9B

๐Ÿ‡จ๐Ÿ‡ณ Alibaba

Sep 2024 10.2B
Multimodal Open (Restricted)
DS

Galaxy (ๆ˜Ÿๆฑ‰Galaxyๅคงๆจกๅž‹)

๐Ÿ‡จ๐Ÿ‡ณ Dahan Software / Hanweb

Sep 2024 32B
Multimodal Proprietary
AI

LLaVA-MORE 8B

๐Ÿ‡ฎ๐Ÿ‡น AImageLab

Sep 2024 8B
Multimodal Open Weights
GR

GameNGen

๐Ÿ‡บ๐Ÿ‡ธ Google Research

Aug 2024
Multimodal Proprietary
Microsoft

Kosmos-2.5

๐Ÿ‡บ๐Ÿ‡ธ Microsoft

Aug 2024 1.3B
Multimodal Open (Restricted)
NVIDIA

LongVILA-7B

๐Ÿ‡บ๐Ÿ‡ธ NVIDIA

Aug 2024 7B
Multimodal Open Weights
SR

xGen-MM (BLIP-3)

๐Ÿ‡บ๐Ÿ‡ธ Salesforce Research

Aug 2024 4B
Multimodal Proprietary
IN

Zhiye LLM (ๆตชๆฝฎ็Ÿฅไธšๅคงๆจกๅž‹)

๐Ÿ‡จ๐Ÿ‡ณ Inspur

Aug 2024
Multimodal Proprietary
OpenAI

GPT-4o (Aug 2024)

๐Ÿ‡บ๐Ÿ‡ธ OpenAI

Aug 2024
Multimodal Proprietary
ByteDance

LLaVA-OV-72B

๐Ÿ‡จ๐Ÿ‡ณ ByteDance

Aug 2024 72B
Multimodal Open (Restricted)
ByteDance

LLaVA-OV-7B

๐Ÿ‡จ๐Ÿ‡ณ ByteDance

Aug 2024 7.6B
Multimodal Open (Restricted)
CT

Xingchen Multimodal Model (ไธญ็”ตไฟกไบบๅทฅๆ™บ่ƒฝ็ง‘ๆŠ€๏ผˆๅŒ—ไบฌ๏ผ‰ๆœ‰้™ๅ…ฌๅธ)

๐Ÿ‡จ๐Ÿ‡ณ China Telecom

Jul 2024
Multimodal Proprietary
LL

LLaVA-NeXT-32B-Qwen

LMMs-Lab

Jul 2024 32B
Multimodal Open (Restricted)
Qwen

LLaVA-NeXT-32B-Qwen

๐Ÿ‡จ๐Ÿ‡ณ Qwen

Jul 2024 32B
Multimodal Open (Restricted)
SA

InternVL2-Llama3-76B

๐Ÿ‡จ๐Ÿ‡ณ Shanghai AI Lab

Jul 2024 76B
Multimodal Open (Restricted)
Qwen

Qwen2-Audio

๐Ÿ‡จ๐Ÿ‡ณ Qwen

Jul 2024 8.2B
Multimodal Open (Restricted)
SE

SenseChat 5.5

๐Ÿ‡จ๐Ÿ‡ณ SenseTime

Jul 2024 600B
Multimodal Proprietary
SA

InternVL2 26B

๐Ÿ‡จ๐Ÿ‡ณ Shanghai AI Lab

Jul 2024 25.5B
Multimodal Open (Restricted)
SA

InternVL2-40B

๐Ÿ‡จ๐Ÿ‡ณ Shanghai AI Lab

Jul 2024 40.1B
Multimodal Open (Restricted)
BA

Ernie 4.0 Turbo

๐Ÿ‡จ๐Ÿ‡ณ Baidu

Jun 2024
Multimodal Proprietary
NY

Cambrian-1-13B

๐Ÿ‡บ๐Ÿ‡ธ New York University (NYU)

Jun 2024 13B
Multimodal Open (Restricted)

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.