Docs · Questions
StreamDiffusionTDUsage
What are all the StreamDiffusionTD release versions and their release dates?
StreamDiffusionTD Release History
StreamDiffusionTD has had 20 major releases from January 2024 to November 2026. Here’s the complete version history in reverse chronological order (newest first):
Recent Releases (2025-2026)
v0.4.2 - November 9, 2026
RunPod cloud mode on your own RunPod account, replacing Daydream, with a GPU menu, Output Smoothing and sign-in that renews itself. Install Step 1 updates an install in place. CUDA-Link 1.14.1, SDXL build memory, Hugging Face offline mode opt-in.
v0.4.1 - September 21, 2026
Maintenance release on v0.4.0: install lookup and recovery, updater fixes, ControlNet preprocessing, pose engine auto-build, NaN recovery.
v0.4.0 - August 22, 2026
ControlNet, IP Adapter, StreamV2V and live LoRA weight together on TensorRT. GPU to GPU transport (cuda-link), TensorRT 10.16, torch 2.11, CUDA 12.8 only.
v0.3.1 - March 24, 2026
Stability and feature update. Built-in FX processors (feedback loop, color grading, latent drift, chromatic aberration), StreamV2V cached attention is back (TRT required), new installer CLI with verify/diagnose/repair, depth TRT auto-build, auto-updates, scheduler/sampler selection, dozens of bug fixes.
v0.3.0 - November 6, 2025
Major rebuild with SDXL + IP Adapter + TensorRT support. No installation required for hosted backend.
v0.2.99 - August 21, 2025
Built-in cloud backend (Daydream API), 5xxx GPU support, Mac improvements, multi-controlnet + TensorRT for cloud version, feedback loop toggle.
v0.2.6 - February 12, 2025
Greatly improved Windows installation, Mac (Apple Silicon) support with built-in installation, Textual Inversion/Embeddings support, multiple LoRA loading fixes, Easy OP Loading feature.
2024 Releases
v0.2.3 - October 8, 2024
SDXL ControlNet support (openpose, canny, scribble, depth), installation bug fixes, model dropdown improvements, FPS counter.
v0.2.2 - September 5, 2024
Resolution switching bug fix, improved local model support, LoRA weight adjustment improvements, install check labels.
v0.2.0 - August 18, 2024
Major update: ControlNet support, V2V temporal consistency, Pause/Play/Unload commands, offline usage, LoRA loading improvements.
v0.1.11 - May 16, 2024
Mac (Apple Silicon) support, Model Preset Config system, non-512x512 TensorRT resolutions.
v0.1.9 - May 3, 2024
Bug fixes for resolution switching and shared memory.
v0.1.8 - April 21, 2024
Major performance update: Replaced NDI+Spout with direct memory buffer for less latency and higher FPS, improved diffusion quality.
v0.1.7 - April 11, 2024
SDXL-Turbo support, SDXS model support, improved LoRA loading, git LFS check.
v0.1.5 - February 22, 2024
Installation guide, Hugging Face model downloader, LoRA dropdown, better 1.5 model handling.
v0.1.4 - February 15, 2024
Config/model error fix.
v0.1.3 - February 14, 2024
Txt2img support, offline model support, “My Models” menu, CUDA 12.1 installation fixes.
v0.1.2 - February 5, 2024
Prompt Blocks, Seed Blocks, improved installation process, Spout BGR fix.
v0.1.1 - January 15, 2024
NDI memory leak fix.
v0.1.0 - January 11, 2024
Initial release: Automated setup, stream control, live parameter updates, LoRA support, NDI/Spout integration, callback functions.
Latest Version: v0.4.2 (November 9, 2026)
Total Releases: 20
Development Span: January 2024 - November 2026 (34 months)