Local Ai
LTX 2.3 - Image + Audio + Video ControlNet (IC-LoRA) to Video
LTX-2.3 is a DiT-based audio-video foundation model from Lightricks that generates synchronized video and audio within a single model pass, representing a significant upgrade over LTX-2 with improv...
LTX-2.3 is a DiT-based audio-video foundation model from Lightricks that generates synchronized video and audio within a single model pass, representing a significant upgrade over LTX-2 with improved audio-visual quality and enhanced prompt adherence. It supports IC-LoRA (In-Context LoRA), which enables fine-grained video-to-video and image-to-video control by conditioning generation on reference frames, images, and audio at inference time — functioning similarly to ControlNet for video. The model ships with open weights, runs locally via ComfyUI with native IC-LoRA node support, and includes specialized IC-LoRA variants for union control, motion-track control, and identity/audio conditioning.
Related
- ControlNet vs LoRA
- LTX 2.3 Lip Sync Music Clip -- Drake - Toosie Slide
- VoxCPM TTS model + LoRa training abilities right in Comfy
- Creating unique visual styles for your videos with Wan 2.1
- Live AI video is doing too much lifting as a term. Here's a breakdown of what people actually mean.
Source: local-ai