ComfyUI Node
H3 Relay · LTX Upscale Model Loader
A ComfyUI node in H3 Relay/loaders with 10 inputs and 1 output.
H3 Relay · LTX Upscale Model Loader
- ltx_model
◄model_nameltx-2.5-22b-dev-transformer-comfy-int8-convrot.safetensors►
◄vae_nameltx-2.5-video-vae-bf16.safetensors►
◄latent_2x_model_nameltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors►
◄text_encoder_namegemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors►
◄distilled_loraltx-2.5-22b-distilled-lora-450-bf16.safetensors►
◄distilled_strength1.00►
◄pixel_upscale_ic_loraltx-2.5-22b-ic-lora-pixel-spatial-upscaler-x2-1.0.safetensors►
◄pixel_upscale_ic_strength1.00►
◄weight_dtypedefault►
◄manual_cache_revisionv1►
CategoryH3 Relay/loaders
Inputs (10)
| Name | Type | Default | Description |
|---|---|---|---|
| model_name | COMBO | ltx-2.5-22b-dev-transformer-comfy-int8-convrot.safetensors | 0 options: |
| vae_name | COMBO | ltx-2.5-video-vae-bf16.safetensors | 0 options: |
| latent_2x_model_name | COMBO | ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors | Learned 2x latent spatial expansion. This establishes the high-resolution target latent before diffusion refinement. |
| text_encoder_name | COMBO | gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors | 0 options: |
| distilled_lora | COMBO | ltx-2.5-22b-distilled-lora-450-bf16.safetensors | Adapts the LTX dev transformer for the fast low-step schedule. It does not perform the spatial upscale. |
| distilled_strength | FLOAT | 1.00-4–4 | — |
| pixel_upscale_ic_lora | COMBO | ltx-2.5-22b-ic-lora-pixel-spatial-upscaler-x2-1.0.safetensors | Reference-conditioned IC-LoRA. It guides diffusion from the original low-resolution pixel video after the latent has been expanded 2x. |
| pixel_upscale_ic_strength | FLOAT | 1.000–2 | — |
| weight_dtype | COMBO | default | 4 options: default, fp8_e4m3fn, fp8_e4m3fn_fast, fp8_e5m2 |
| manual_cache_revision | STRING | v1 | Manual cache reset only. Change this when replacing model contents without changing the filename; normal model, LoRA, dtype, strength, and attention changes are tracked automatically. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| ltx_model | H3_RELAY_MODEL | — |