Features
Face consistency via Person profiles
Style cloning from YouTube URL
Generate 1-4 thumbnail variations per prompt
Template mode for podcast/branding reuse
Inspiration library to save and reuse styles
Pixelate tool to bypass AI policy blocks
Live editor with preview before generation
Automatic version history dashboard
1280x720 PNG output for YouTube
GPT Image 2, Gemini 3 Pro, Gemini 3.1 Flash models
Podcast mode: automatic guest swapping
Tagging (@mentions) to reference persons/styles
Anonymize inspirations for privacy
Built-in YouTube Thumbnail Downloader
Built-in YouTube Video Card Generator
Exceptional prompt understanding for complex scenes
Lifelike photoreal images with human feel
Unmatched vector generation with editable graphics
Quality text generation in images
Consistent styles without training via image drop-in
Custom style creation and reuse
Exploration Mode for iterative generation
API access for developers
Integration with other frontier AI models
Production-ready shots with tasteful composition
Complex multi-character scenes
Agentic mode for conversational creation
Background removal, inpainting, outpainting via API
