TinyShip 1.8.0 Released
This release officially adds AI video generation and completes a key capability in TinyShip's AI feature set.
✨ Core Additions
AI Video Generation Page: Added /video-generate to generate videos from prompts or first-frame images
Multi-Provider Support: Unified integration for multiple video providers
- Volcengine (Seedance)
- Aliyun Wanxiang (Wan)
- FAL (Kling)
Credits Integration: Video generation is integrated with credits and supports per-model fixed costs
Permission Alignment: Video generation API now follows the same auth policy as image generation API (login required by default)
🎯 AI Video Generation Features
Two Input Modes:
- Text to Video
- First Frame Image to Video
Unified Generation Interface: Provider differences are abstracted behind one consistent API and frontend interaction
UX Improvements:
- First-frame image preview after upload
- Default demo video and prompt for quick testing
🔧 Technical Highlights
Provider Configuration Alignment: Default video provider switched to Volcengine with unified provider order and model presentation
FAL Parameter Adaptation: Fixed FAL duration format compatibility to prevent request validation errors
Response Handling Optimization: Prefer hosted video URLs to avoid oversized base64 responses
Credit Consistency Fixes: Unified model resolution and credit calculation logic to avoid default-model billing mismatches
⚠️ Breaking Changes
Input Mode Simplification: Removed the first+last frame mode and kept only text-to-video and first-frame image-to-video.