Skip to main content

Model library

Every media model behind one MCP interface.

Agents can generate images, videos, and audio without learning a new provider API for each model. AgentFramer keeps model constraints, defaults, and pricing metadata in one catalog.

Image Models

Prompt-to-image models with verified dimensions and defaults.

Nano Banana 2

nano-banana-2

Default

Google's flagship generation model for detailed, instruction-following visual creation.

1024x1024

Recraft V4

recraft-v4

Professional-grade model for design and marketing workflows.

1024x1024

Recraft V4 Pro

recraft-v4-pro

Advanced 2K model for brand-critical creative production.

2048x2048

GPT Image 2

gpt-image-2

OpenAI's latest image model via Runware.

2048x2048

Video Models

Text and image-to-video models with duration and aspect-ratio constraints.

Seedance 2.0

seedance-2.0

Default

Multimodal video generation with synchronized audio.

5s, 16:9

Seedance 2.0 Fast

seedance-2.0-fast

Fast multimodal video generation for rapid iteration.

5s, 16:9

Kling Video 3.0 Pro

klingai-video-3.0-pro

High-quality cinematic video generation with native audio.

5s, 16:9

Kling Video 3.0 4K

klingai-video-3.0-4k

4K cinematic video generation with native audio.

5s, 16:9

HappyHorse 1.0

alibaba-happyhorse-1.0

Alibaba video generation with 720p and 1080p output.

5s, 16:9

Grok Imagine Video

grok-imagine-video

xAI cinematic video with prompt-synchronized audio.

6s, 16:9

HeyGen Video Agent

heygen-video-agent

Prompt-only multi-scene avatar and presenter videos.

30s, 16:9

Audio Models

Music, voice, and sound generation models for agentic workflows.

Eleven Music v1

eleven-music-v1

Default

Text-to-music model for multilingual music generation.

30s default

Agents can omit model IDs.

Every tool has a recommended default. Advanced agents can override models when they need a specific provider, duration, size, or output style.