My approach: I generate everything in "Fast" mode first to quickly iterate and find the best results, then I selectively upscale only the videos worth keeping by reusing the same seed.<h2 id="generation-modalities">Generation Modalities</h2><ul><li>📝➡️🎥 Text-to-Video Create completely new videos from scratch using text prompts.</li><li>🖼️➡️🎥 Image-to-Video Animate static reference images using text prompts.</li><li>🖼️🖼️➡️🎥 First-Last-Frame-to-Video Generates a coherent video sequence that bridges a defined starting and ending image.</li><li>🎥➡️🎥 Video-to-Video Generates synchronized audio and speech driven by video visuals. Offers options to pass through or entirely regenerate the source video.</li></ul><h2 id="audio-input-settings">Audio Input Settings</h2><ul><li>🔇 No Audio Input No external audio file used. The AI generates new audio based on your text prompt.</li><li>🔊 Audio Input Upload an existing voice or music file to drive the animation and lip-sync.</li></ul><h2 id="vram-optimization-and-long-videos">🚀 VRAM Optimization &amp; Long Videos</h2><h3 id="chunking-settings-for-longer-videos">⚙️ Chunking Settings for Longer Videos</h3>Adjust ffn_chunks based on your video length (at 24fps):<ul><li>10 seconds (240 frames): ffn_chunks=1-2</li><li>15 seconds (360 frames): ffn_chunks=2-4</li><li>20 seconds (480 frames): ffn_chunks=4-6</li><li>25 seconds (600 frames): ffn_chunks=8-10</li><li>33 seconds (800 frames): ffn_chunks=12-16</li></ul>If you encounter OOM (Out of Memory) errors:<ul><li>Increase ffn_chunks value</li><li>Reduce resolution slightly</li></ul>

LTX2 All-in-One ComfyUI Workflow

ltx2AllInOneComfyui_ltx23DistilledaioV23.zip

LTX2.3 DistilledAIO v2.3

LTX2.3 DistilledAIO v2.4

LTX2 Distilled AIO v2.2

LTX2 Distilled AIO v2.0

LTX2 Distilled AIO v1.3

LTX2 Distilled AIO v1.2

LTX2 Distilled AIO v1.0

LTX2 Distilled AIO v0.95

LTX2 Distilled AIO v0.9

Files with SHA256: e0df37399d78756913e02dfa65dde70ab7dc3674fa5efb2ae5889741e26a4346