New Models, Quick Voice Creation & Smoother Canvas
· Update
🚀 New Models
- Seedream 5.5, Image 2.6, and Image 2.5 are now live on Pixmax. Image-series input limits are updated (max 30 MB per image, 100 MB per request).
🎙️ Quick Voice Creation (Seed Audio 1.0)
- Pixmax now supports one-click voice creation.
- In Canvas, easily box-select a character card and voice to save them to the Element Library at once. Image nodes also support one-click voice creation.
- Element Library can now intelligently match audio to images, streamlining setup for projects with many background characters.
🌍 Creation Controls
- Added image and video watermark toggles.
- Doodle text now supports adjustable font size and color palette selection.
✨ Canvas Improvements
- Added presets for one-click character three-view and headshot extraction, with custom prompt support.
- Inpainting results can now be seamlessly merged back onto the original image.
- Assets from the Asset Library can be dragged directly onto the Canvas. Multiple assets auto-arrange in blank areas, and the library now supports grid/list view toggling.
- History previews display the full image or a video preview above each result.
- New nodes for Upscale, Edit, or Outpaint tasks no longer overlap the source node and feature clearer naming conventions (e.g., "Upscale xxx – Upscale").
- Single-node task history now features thumbnail navigation and direct insertion into the Canvas.