StreamDiffusionTD

Key Features

• Real-time Generation: StreamDiffusion img2img / controlnet / v2v
• Local & Cloud: Run locally or use Daydream cloud backend
• ControlNet: Use in2 for ControlNet (Daydream uses in1)
• V2V Temporal Consistency: Smoother video-to-video output
• TensorRT Acceleration: Even faster performance on supported GPUs

Quick Start

1. Install StreamDiffusionTD - Setup for local or cloud
2. Usage Examples - Essential workflows
3. Key Parameters - Core settings reference

System Requirements

Local Mode

• OS: Windows 10/11, macOS (Apple Silicon)
• GPU: NVIDIA with 6GB+ VRAM (RTX 3060+)
• CUDA: 12.1 recommended (12.x supported, 5090/5080 with 12.8)
• Python: 3.11.9

Daydream Cloud Mode

• Python: 3.11.9
• FFmpeg: Required for video processing

Version Information

Current Version: 0.2.99

• Daydream cloud backend support
• RTX 5090/5080 compatibility (limited)
• Enhanced Mac installation
• Feedback loop toggle

Download

Available on Patreon for supporters.