SeeDream 4.5 - ByteDance's Latest Image Model
- Upgraded to SeeDream 4.5 - ByteDance's enhanced image generation model
- Up to 14 reference images for image-to-image generation
- Sequential generation mode - generate up to 15 related images automatically
- New 21:9 ultrawide aspect ratio support
- Automatic input image dimension matching
- Enhanced image quality and detail rendering
- 2K and 4K resolution options (removed 1K)
- Full aspect ratio selector with 9 options
Veo 3.1 Enhanced - Full API Support
- Duration control for Veo 3.1 - choose between 4, 6, or 8 second videos
- Generate Audio toggle - enable or disable AI-generated synchronized audio
- Last Frame selector - create smooth transitions between start and end frames
- Reference Images support - up to 3 images for subject-consistent video generation (R2V)
- Streamlined video model selection - now featuring Veo 3.1 and Sora 2
- Removed legacy Veo 3 and Veo 3 Fast models (superseded by Veo 3.1)
- Cleaner UI with model-specific controls
Nano Banana Pro - Google's Latest Image Generation
- Added Nano Banana Pro - Google's state-of-the-art image generation model based on Gemini 3 Pro
- Text rendering support - create images with legible text in multiple languages
- Up to 14 reference images for complex transformations and multi-person consistency
- Resolution options: 1K, 2K, and 4K output sizes
- Available in both Text-to-Image and Image Reference modes
- Image Reference tab now supports multi-image selection for supported models
- Improved model switching with context-aware settings
Flux 2 Pro & Performance Improvements
- Upgraded to Flux 2 Pro - Black Forest Labs' latest and most advanced image generation model
- New resolution options for Flux 2 Pro: 0.5 MP, 1 MP, 2 MP, and 4 MP output sizes
- Enhanced image quality with improved detail and consistency
- Videos in gallery now use hover-to-preview - hover to play, no thumbnails needed
- Streamlined video processing for faster gallery loading
- Fixed character loss when typing quickly in prompt input fields
- Fixed prompt input lag on AI Image, Video, and Audio generation pages
- Added debouncing to all prompt inputs for smoother typing experience
- Fixed video play button positioning in gallery views