LTX-2 Text & Image to Video - ComfyUI & One Click Installer

The new LTX-2 19B model pushes the boundaries of local AI video generation, combining lightning-fast speed, low VRAM usage, and stunning cinematic quality. With this release, you can now generate full motion clips from simple text prompts or static images — all running directly on your PC with ComfyUI.
This post includes two powerful workflows — Text-to-Video and Image-to-Video — each optimized for low VRAM systems and minimal setup. To make things even easier, the one-click Windows installer installs ComfyUI, all dependencies, and preloads every essential model automatically.
 
Preloaded Models Within the Installer (Low VRAM)
ltx-2-19b-embeddings_connector_dev_bf16.safetensors (clip folder) — Hugging Face
gemma-3-12b-it-UD-Q4_K_XL.gguf (clip folder) — Hugging Face
LTX2_audio_vae_bf16.safetensors (vae folder) — Hugging Face
LTX2_video_vae_bf16.safetensors (vae folder) — Hugging Face
LTX-2-dev-Q3_K_S.gguf (unet folder) — Hugging Face
ltx-2-spatial-upscaler-x2-1.0.safetensors upscale model — Hugging Face
ltx-2-temporal-upscaler-x2-1.0.safetensors upscale model — Hugging Face
ltx-2-19b-ic-lora-detailer.safetensors LoRA — Hugging Face
ltx-2-19b-distilled-lora-384.safetensors LoRA — Hugging Face
ltx-2-19b-lora-camera-control-static.safetensors LoRA — Hugging Face
MelBandRoformer_fp32.safetensors — Hugging Face
LTX-2 19B standard diffusion models: Hugging Face
 
Speed
Generate 10-second 480p text-to-video clips in under a minute on an RTX 4090 (24 GB VRAM), and 10-second 480p image-to-video clips in under three minutes. The workflow dynamically adapts to your hardware, allowing 720p and high frame-rate output (24fps+) on stronger GPUs.
 
System Requirements
Nvidia RTX 30XX / 40XX / 50XX GPU (FP16 supported)
CUDA-compatible GPU (minimum 8 GB VRAM, 12 GB+ recommended)
Windows OS
Minimum 50 GB free storage
 
What's Included
Portable ComfyUI Windows Installer, fully configured for LTX-2 image-to-video and text-to-video
Automated downloads for all models and nodes
Easy-to-use workflows for both new and advanced users
Fast Groups Bypasser support for quick workflow toggles
 
Usage Notes
Works with both LTX-2 19B GGUF and the standard LTX-2 19B diffusion models.
Toggle workflow layers via the Fast Groups Bypasser (left sidebar).
Use rich, detailed prompts for smoother motion and narrative coherence.
For image-to-video, always start with high-resolution source images.
To customize workflow settings, click the arrow at the top-right of the subgraph prompt node.