The sheet comes through
Limit, from LTX 2.5 Creative Lab LoRAs, tested locally


- LoRAs
- Ingredients at 1.4 and Slow Motion Control at 1.0
- Output
- 768×448, 121 frames at 24fps
- Sampling
- 8 steps, euler_ancestral_cfg_pp
- Seed
- 42
- Generated in
- 3 min
- Peak VRAM
- 11.0 GiB
- Ran on
- RTX 3060, 12GB (local)
What was wired in
Reference sheet
Rooftop sheetIngredients sheet: character, camera and rooftop panels on blackDownload PNG0.7 MB
The sheet is clean and evenly spaced with one character on it, and the output still reproduces the sheet panel by panel. When this happened, a small prompt change and a new seed usually cleared it.
Reference sheet: Top Row Left (Character): a portrait photo of a young woman with shoulder-length dark hair streaked pink and blue, large gold hoop earrings, a small leaf tattoo on her neck and a beige knitted sweater. Top Row Right (Prop): a chunky, boxy bright-yellow camera with a round lens and a textured grille on its side. Bottom Row Left (Setting): an empty rooftop terrace with a brick parapet, potted olive trees, string lights and a city skyline. Generated video: On the rooftop terrace at golden hour, the young woman with pink-and-blue streaked hair and gold hoop earrings, wearing the beige knitted sweater, faces the lens holding the chunky bright-yellow camera up to her right eye with both hands. She presses the shutter, then slowly lowers the camera to her chest with both hands still on it, and looks straight into the lens with a small smile as a breeze lifts her streaked hair. Static medium shot from the waist up, the brick parapet and olive trees behind her. The sound is the click of the shutter, a light breeze and distant traffic.
worst quality, inconsistent motion, blurry, jittery, distorted
Notes
This run also had the Slow Motion LoRA chained in. The same failure shows up in the chaining tests at the end of the page.
Run it yourself
Download everything for this test (.zip, 0.7 MB). It holds the workflow, every input it loads and a README. Copy the inputs into ComfyUI/input, put the models below in place and drag the workflow onto ComfyUI.
| File | Put it in | Size |
|---|---|---|
| ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors gated | models/diffusion_models | 21.5 GB |
| gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors gated | models/text_encoders | 15.4 GB |
| ltx-2.5-video-vae-bf16.safetensors gated | models/vae | 1.5 GB |
| ltx-2.5-audio-vae-bf16.safetensors gated | models/vae | 364.9 MB |
| ltx-2.5-22b-ic-lora-ingredients-0.9.safetensors gated | models/loras | 1.3 GB |
| ltx-2.5-22b-lora-slow-motion-control-1.0.safetensors gated | models/loras | 654.5 MB |
The Lightricks repositories (the LTX 2.5 base model and the LoRAs) are gated. Log in to Hugging Face and accept the terms on each repository page before downloading.
Custom nodes: ComfyUI-LTXVideo at commit 3bf3ca6. Everything else is part of ComfyUI 0.37.0 (commit 0f74f7fb, frontend 1.53.6).