An end-to-end video streaming middleware for Express.js with a modern custom web video player powered by HLS and Media Source Extensions.
🚧 StreamForge is under active development. While v0.3 is complete, APIs may continue to evolve before the first stable release.
StreamForge automates the complete video streaming pipeline—from ingest to playback—as a plug-and-play Express.js middleware. It generates HLS streams on demand and ships with SFPlayer, a lightweight custom web component built on the Media Source Extensions (MSE) API.
- 🎬 Automatic HLS generation
- 📦 Fragmented MP4 (fMP4) streaming
- ⚙️ On-demand transcoding
- 🎥 Custom
<sf-player>web component - ⏩ Accurate seeking
- 📊 Playback & buffer progress
- 🎚️ Manual quality selection
- ⚡ Retry failed segment downloads
- ⏳ Request timeout handling
- 📡 Live transcoding progress via Server-Sent Events (SSE)
- 🖼️ Picture-in-Picture support
- ⏩ Playback speed controls
- 🧹 Automatic cache cleanup
- 🛡️ Structured error handling with custom error classes
- HLS Generation
- Media Engine
- Segment Loader
- Basic Seeking
- Progress Bar
- Error Handling
- Retry Failed Segments
- Buffer Manager
- Manual Quality Selection
- Playback Speed
- Picture-in-Picture
- Request Timeout
- AbortController Support
- Generation Progress UI
- Buffer Progress Indicator
- Improved Network Layer
- Improved Error Handling
- Stability & Cleanup
- Rich StreamForge Manifest
- Better metadata handling
- Browser capability detection
- Native HLS support
- Automatic playback backend selection
- Unified playback architecture
- Public Player API
- Event API
- Real-time encoding progress
- Better middleware integration
- Improved developer experience
- Cache validation
- Automatic regeneration
- Better retry system
- Improved error reporting
- Adaptive Bitrate (ABR)
- CDN Integration
- Cloud Storage Support
- Player Themes
- Headless Player API
- Subtitles & Captions
- Thumbnail Preview
- Multi-audio Support
- Analytics & QoE Metrics
Contributions, issues, and feature requests are welcome. StreamForge is still evolving, and feedback is greatly appreciated.
MIT