LTX 2.3

LTX 2.3 is Lightricks' open-source audio-visual video generation model. It generates synchronized video and audio in a single pass, supports native vertical output up to 4K, and runs at up to 50 fps.

2.3 ships a rebuilt variational autoencoder, a quadrupled text encoder built on Gemma 3 12B, and a gated-attention text connector for better prompt adherence. Resolution, frame rate, and audio quality all step up over LTX-2.

Yes, natively. The model generates 9:16 directly at full resolution rather than cropping a horizontal frame, which keeps detail and composition right for vertical-first surfaces.

Feed in an audio clip and LTX 2.3 generates matching visuals timed to the audio. Useful for music-driven content, ads keyed to a track, or dialogue-led shorts where the audio sets the pacing.

Open Copilot, describe the shot or attach an audio track, and pick LTX 2.3. For vertical short-form, audio-led content, or anything that benefits from native sound generation, LTX 2.3 is the right pick.

Yes. Lightricks publishes the weights on Hugging Face under the LTX-2 Community License. On Morphic you use the hosted version, no infrastructure to manage, but the underlying model is openly available for local or on-prem use.