🎨 Multimodal AI Models

📊 Modality Breakdown

🏆 Flagship Multimodal Models

The most capable multimodal models across all providers:

Model	Provider	Context	Input	Output	Tool Call	Price (in/out per 1M)
gpt-4o	OpenAI	128K	text, image	text	✓	$2.50/$10
gpt-4.1	OpenAI	1M	text, image	text	✓	$2/$8
claude-sonnet-4	Anthropic	200K	text, image	text	✓	$3/$15
gemini-2.5-pro	Google	1M	text, image, audio, video	text	✓	$1.25/$10
gemini-2.5-flash	Google	1M	text, image, audio, video	text	✓	$0.15/$0.60
llama-4-maverick	Meta	1M	text, image	text	✓	Varies
qwen3-235b-a22b	Alibaba	128K	text, image	text	✓	Varies

👁️ Vision Models (Image Input)

1,487 models can accept images as input alongside text. These are the most common type of multimodal model:

→ See all 1,487 vision models compared

🎤 Audio Input Models

118 models can process audio input — for transcription, voice analysis, and audio understanding:

Model	Provider	Audio Capabilities	Context
gemini-2.5-pro	Google	Audio understanding + transcription	1M
gemini-2.5-flash	Google	Audio understanding + transcription	1M
gpt-4o-audio-preview	OpenAI	Audio input + output	128K
claude-sonnet-4	Anthropic	Audio transcription	200K

🖼️ Image Generation Models

28 models can generate images from text descriptions. This is a rapidly growing category:

Model	Provider	Capabilities
gpt-image-1	OpenAI	Text-to-image, image editing
dall-e-3	OpenAI	Text-to-image generation
flux-1.1-pro	Black Forest Labs	High-quality text-to-image
stable-diffusion-3.5	Stability AI	Open-weight text-to-image

→ See all 28 image generation models

🔊 Audio Output Models

34 models can generate audio output — for text-to-speech, voice cloning, and audio generation:

🎬 Video Understanding Models

167 models can process video input — for video analysis, summarization, and content understanding:

🤔 Choosing the Right Multimodal Model

🔗 Related Comparisons