Saltar al contenido
WorkflowPublicadoExterno
Abrir fuente

LTX 2 VIDEO DETAILER — v1.0

V1.0

FamiliaLTX-VideoVersiónVRAM estimada16 GiBBcalidad

Perfil de ejecución

Acceso
Flujo externo
Estado
Publicado
Versión
LTX 2 DETAILER 10 SECOND.json
Fuente
Civitai
Actualizado
18 oct 2026

What you need

GPU memory
16 GiB
Model files
33.9 GiB known subtotal

Remaining hardware or download-size estimates are not available yet.

Requisitos

Modelos

7

Paquetes de nodos

7
  • ComfyLiterals

    Encontrado

    Paquete de nodos · Registry

  • comfyui-easy-use

    Encontrado

    Paquete de nodos · Registry

  • comfyui-florence2

    Posible

    Paquete de nodos · Registry

  • ComfyUI-GGUF

    Posible

    Paquete de nodos · Registry

  • ComfyUI-LTXVideo

    Encontrado

    Paquete de nodos · Registry

  • comfyui-videohelpersuite

    Encontrado

    Paquete de nodos · Registry

  • RES4LYF

    Encontrado

    Paquete de nodos · Registry

Comentario generado por IA

Explicación generada por IA basada en los detalles de la fuente y la configuración. Las sugerencias están claramente identificadas.

Propósito

  • LTX 2 VIDEO DETAILER is a ComfyUI workflow for detailing 5- or 10-second 1280 × 720 videos and supporting temporal consistency; it takes a video and produces a detailed video without changing the resolution.

Flujo

  • Use it in ComfyUI as an LTXV workflow from AITold, version v1.0.

Entradas y salidas

  • Choose the 5-second workflow for a 5-second input video or the 10-second workflow for a 10-second input video.

  • The described format is 1280 × 720 video, and the output keeps the input resolution.

Configuración inicial

  • Setup uses a different LTX installation process with quantized models, the Vantage GGUF Unet Loader from VantageWithAI, Vantage-Nodes, `pip install -r requirements.txt`, and a ComfyUI restart.

  • Before setup, verify the exact package or repository for ComfyUI-GGUF and comfyui-florence2, and confirm the intended LTX model files.

    Sugerencia · sin verificar

Dependencias

  • The workflow lists these node packs: comfyui-videohelpersuite, RES4LYF, comfyui-easy-use, ComfyLiterals, comfyui-florence2, ComfyUI-LTXVideo, and ComfyUI-GGUF.

  • The setup names these model files: gemma_3_12B_it_fp8_e4m3fn.safetensors, ltx-2-19b-distilled-lora-384.safetensors, ltx-2-19b-ic-lora-detailer.safetensors, ltx-2-19b-embeddings_connector_dev_bf16.safetensors, and ltx-2-19b-dev_Q4_K_M.gguf.

  • It also names LTX2_video_vae_bf16.safetensors, LTX2_video_vaeold_bf16.safetensors, and microsoft/Florence-2-large.

Configuración

  • The setup suggests a dedicated LTX-2 launch file with `--lowvram --disable-xformers --use-pytorch-cross-attention --reserve-vram 2 --disable-smart-memory`.

    Sugerencia · sin verificar

  • For a 10-second video on the 5-second workflow, the instructions suggest 32+ GB RAM; if the machine struggles, they suggest lowering the resolution to 480p.

    Sugerencia · sin verificar

Precauciones

  • The workflow is described as a test workflow; artifacts may still appear, and the stated hardware test used 16 GB VRAM.

  • If repeated runs produce black or red grids, the instructions attribute this to VRAM throttling and advise restarting the PC; switching between the 5- and 10-second workflows may require closing and reopening ComfyUI because their VAE encoding methods and cache differ.

    Sugerencia · sin verificar

¿Es necesario editarlo?

Inicia sesión para solicitar una edición.

Fuentes

1 fuente
CivitaiPrincipalVersión 2600627
civitai.com/models/2311504

Verificado 18 oct

Extractos de fuentes

3 extractos

Source context: 953 downloads · Type Workflows · Base model LTXV

Source context: 894 downloads · Type Workflows · Base model LTXV

Source context: 865 downloads · Type Workflows · Base model LTXV