NexusBTA v0.2.22 launches web UI with pre-made ComfyUI workflows
New UI bundles ComfyUI workflows for 15+ models including FLUX, WAN 2.2, and Trellis 2.
NexusBTA v0.2.22 is now available, offering a web UI powered by ComfyUI that comes with pre-made workflows for over a dozen popular models. The new release supports generation models including ANIMA, WAN 2.2, LTX 2.3, Stable Diffusion 1.5 and XL, Illustrious, Pony, FLUX (including FLUX 2 Klein and FLUX Dev), Qwen Image Edit, Z Image (standard and Turbo), Lumina, and Trellis 2 for 3D model generation. Users can simply run the provided .bat file to start generating, making it a drop-in solution for anyone wanting to explore different AI art and video pipelines without manually wiring up ComfyUI workflows.
This update brings major workflow and runtime improvements. Highlights include better Qwen and FLUX inpainting with improved multiview routing, updated start/end frame and loop workflows for WAN 2.2 and LTX 2.3, and enhanced motion transfer handling. Several fixes address Civitai modal issues, model path scanning, and runtime/bootstrap reliability for startup, custom nodes, RunPod/Docker, LAN, and tunnel usage. The extras section also received updates for RTX Super Resolution, PiD upscale, FlashVSR, and SeedVR2, along with better dependency and model-path handling. Full release notes are available on GitHub.
- Pre-made workflows support 15+ models: ANIMA, WAN 2.2, LTX 2.3, SD 1.5/XL, Illustrious, Pony, FLUX, Qwen, Z Image, Lumina, Trellis 2.
- Key improvements: better Qwen/Flux inpaint, multiview routing, WAN/LTX start/end frame and loop workflows, motion transfer fixes.
- Runtime fixes for startup, custom nodes, RunPod/Docker, LAN, and tunnel; extras updated include RTX Super Resolution, PiD upscale, FlashVSR, SeedVR2.
Why It Matters
Streamlines AI image/video generation with one-click workflows, broad model compatibility, and critical runtime stability fixes.