ComfyUI Node: Halo-TeaCache v1.0.0 (AMD LTX2)
Run ComfyUI workflows without the setup
No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.
Category
Halo-TeaCache
Inputs
model MODEL
rel_l1_thresh FLOAT
start_percent FLOAT
end_percent FLOAT
Outputs
MODEL
Extension: Halo-TeaCache
AMD Unified Memory TeaCache for LTX2 (LTXAV). Lean single-file caching for 48-layer dual-stream transformer. Caches both video and audio residuals. No CPU offload - designed for AMD APUs with unified memory. ~35% speedup on cached denoising steps.
Authored by bkpaine1
Looking for a different node?
Run ComfyUI workflows without the setup
No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.