- Add draft mode toggle for generating 480p preview videos at reduced cost
- Add draft_task_id input/output for rendering full resolution from draft
- Implement OUTPUT_NODE pattern to send task ID to frontend via ui data
- Add JavaScript listener to auto-populate draft_task_id after draft generation
- Add submit_and_get_result_with_task_id to API handler for returning task IDs
- Support draft parameter in video generation API calls
Draft workflow:
1. Enable draft_mode=true to generate 480p preview
2. Task ID auto-fills after completion
3. Set draft_mode=false to render full resolution using stored draft
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add Seedream 4 unified node combining text-to-image, image-to-image, and multi-image blending
- Implement dynamic image inputs that automatically add/remove slots as you connect (up to 10 images)
- Add three batch generation modes:
- Single: Generate one image with optional seed control
- Parallel: Generate multiple independent variations with concurrent API calls
- Sequential: Generate thematically related images using sequential_image_generation API
- Add Seedance Pro fast model option (seedance-1-0-pro-fast-251015)
- Fast model automatically disables end_frame support (not supported by API)
- Update README with comprehensive Seedream 4 usage guide and Seedance Pro model selection
- Add optional end_frame parameter to Seedance Pro image-to-video mode
- Automatically detect and use first+last frame generation when end frame is provided
- Update documentation to reflect new functionality
- Maintain backward compatibility with existing workflows
- Prevent console spam from base64 image data in API payloads
- Recursively truncate base64 strings in dictionaries and lists
- Show first 50 chars + truncation message for debugging clarity
- Update title from ComfyUI BytePlus API to ComfyUI Seed API Integration
- Rebrand all node categories and descriptions to use Seed naming
- Add documentation for new Seed 1.6 Chat node with session memory
- Update installation instructions and repository URLs
- Maintain API key configuration documentation for BytePlus endpoints
- Change package name from comfyui-byteplus-api to comfyui-seed-api
- Update description to include video generation, image generation, and chat models
- Update author and repository URLs to reflect Seed branding
- Update class names to SeedreamTextToImageNode and SeedEditImageToImageNode
- Change category to Seed/ImageGeneration
- Rename utility classes to SeedResultProcessor and SeedImageApiHandler
- Update all error messages and display names for consistent Seed branding
- Update class names and categories to use Seed branding
- Change category to Seed/VideoGeneration
- Update error messages and display names consistently
- Maintain all functionality with new Seed-focused naming
- Add BytePlusChatApiHandler for chat completions with thinking mode and reasoning effort
- Add BytePlusChatUtils for message formatting (text and multimodal)
- Implement response parsing and error handling for chat endpoints
- Support for conversation history and session management
- Implements consolidated chat node supporting both seed-1-6-250615 and seed-1-6-flash-250715 models
- Features session-based conversation memory for continuous interactions
- Supports multimodal input (up to 4 images) with configurable detail levels
- Includes thinking mode and reasoning effort controls
- Provides external history import and conversation export
- Default session memory disabled for casual use, opt-in for persistent conversations
