LTX-2.5: ComfyUI Workflow Examples

• Text-to-Video (T2V): Generate videos from text prompts
• Image-to-Video (I2V): Generate videos from an input image
• FLF2V: Interpolate between a first and last frame image

​

Key features

• Diffusion Fidelity Rendering: keyframes-first generation that builds every scene on a grid of high-fidelity keyframes
• Diffusion Video Decoder: sharper faces, legible text, and fewer smears in fast motion
• Native multishot: one generation yields multiple connected shots holding character, environment, lighting, voice, and style across cuts
• Gemma 4 12B text encoder: retains multiple subjects, actions, lighting, detail, and camera direction across complex prompts
• Prompt enhancer: a lightweight model that expands short prompts into detailed instructions
• Auto duration: predicts the right clip length from the described action before diffusion begins
• Native 4K HDR: high-resolution HDR output built for professional finishing, with synchronized audio-video at up to 50 FPS

​

Getting started

1. Update ComfyUI to the latest version
2. Go to Template Library > Video > choose any LTX-2.5 workflow
3. Request access to the model repository on Hugging Face (see below), then follow the pop-up to download models and run the workflow

​

ComfyUI Native Workflows

​

LTX-2.5 Text to Video (T2V)

​

Model downloads

​

Model storage

ComfyUI/
├── 📂 models/
│   ├── 📂 diffusion_models/
│   │   └── ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors
│   ├── 📂 text_encoders/
│   │   ├── gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors
│   │   └── gemma4_e2b_it_bf16.safetensors
│   ├── 📂 vae/
│   │   ├── ltx-2.5-video-vae-bf16.safetensors
│   │   └── ltx-2.5-audio-vae-bf16.safetensors
│   └── 📂 latent_upscale_models/
│       └── ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors

​

Prompting tips

1. Describe the whole scene: Include the shot type, scene, action, characters, and camera movement in one flowing paragraph
2. Audio: Describe sounds and dialogue needed for the scene; the model generates synchronized audio together with the video
3. Use the prompt enhancer: The workflow expands short prompts into detailed cinematic instructions automatically; write a simple scene idea and let the enhancer handle the rest

​

LTX-2.5 Image to Video (I2V)

​

Model downloads

​

Model storage

ComfyUI/
├── 📂 models/
│   ├── 📂 diffusion_models/
│   │   └── ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors
│   ├── 📂 text_encoders/
│   │   ├── gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors
│   │   └── gemma4_e2b_it_bf16.safetensors
│   ├── 📂 vae/
│   │   ├── ltx-2.5-video-vae-bf16.safetensors
│   │   └── ltx-2.5-audio-vae-bf16.safetensors
│   └── 📂 latent_upscale_models/
│       └── ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors

​

Prompting tips

1. Describe what happens next: Write the motion, camera movement, and sounds that follow from the input image; do not re-describe what is already visible
2. Anchor the first frame: Use phrasing like “Use the provided start image as the first frame” when writing a continuation
3. Audio: Describe dialogue and sounds for the scene; the model generates synchronized audio

​

LTX-2.5 FLF2V

​

Model downloads

​

Model storage

ComfyUI/
├── 📂 models/
│   ├── 📂 diffusion_models/
│   │   └── ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors
│   ├── 📂 text_encoders/
│   │   ├── gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors
│   │   └── gemma4_e2b_it_bf16.safetensors
│   └── 📂 vae/
│       ├── ltx-2.5-video-vae-bf16.safetensors
│       └── ltx-2.5-audio-vae-bf16.safetensors

​

Prompting tips

1. Describe the transition: Write what happens between the two frames, including camera movement and audio
2. Keep frames aligned: Use images with the same aspect ratio for a smooth interpolation
3. Audio: Describe dialogue and sounds for the scene; the model generates synchronized audio

​

Resources

• Hugging Face Model
• LTX Model Page
• GitHub Repository
• ComfyUI-LTXVideo