March 2026

Default AI drawing model upgraded again, to Seedream 5.0, for better results
Added Gemini 3.1 Flash Lite as the default text model, for faster AI replies
Integrated the Midjourney v7 art style model, with browsable style categories
Admin dashboard now supports switching between Chinese and English; all popups and dropdowns unified into a frosted-glass style
Added undo/redo to the canvas toolbar
Redesigned the team management page; large file uploads are no longer restricted
Knowledge base search is live
Video compositor officially launched: trim clips, color grade, and add transitions, fully linked with the main canvas
Asset library launched: manage and reuse everything you've generated in one place
More professional 3D lighting controls, with adjustable light angle and color temperature
Team analytics page now supports multiple languages; default drawing model switched to Nano Banana 2
Added an image generation config panel for precise control over reference images and aspect ratio
Added professional color grading, with more precise AI lighting
Project covers are now auto-beautified with an AI-generated image
The prompt library supports drag-and-drop organization into folders
Added a shot configuration assistant for planning video camera work
Unified and polished the AI Q&A interaction experience
Added new payment methods; the Skill Store launched, letting you install new skills for the AI assistant on demand