Brand identity management with colors, fonts, and reference images
Two-model pipeline with draft and refinement
Mask editor for iterative inpainting
Multi-CLI support (Claude Code, Cursor, Windsurf, and MCP-compatible agents)
Token-based consumption tracking and usage dashboard
Secure asset storage with signed CDN URLs
Text-to-video and image-to-video generation with multiple AI models
Outline-to-storyboard workflow for course videos
Reference-image consistency for product videos
Public gallery with prompt copying and remixing
Image generation and editing tools
REST API with webhook support and automation
Multiple premium AI video models including Seedance 2.5, Kling v3 Omni, and Veo 3.1
Free guest trial models with 3 watermarked previews per day
PWA installation for iOS and Android
Support for English and Chinese languages
What makes it different
Persistent visual taste for coding agents ensures every image matches brand automatically
Five core primitives eliminate design decisions from the generation process
Draft-and-refine pipeline delivers photorealistic quality at lower cost
Works with any MCP-compatible coding agent with one subscription
Specialized for two specific jobs: turning teaching outlines into consistent course video series and converting product photos into SKU-consistent shoppable videos
Reference-image consistency ensures generated product videos maintain the same SKU appearance
Multiple AI models in one workflow with production evidence recording
Community gallery with prompt remixing and visual assets for workflow support
REST API for integrating generation workflows into customer systems
AgentBrush
imgmov