Skip to content
Announcements

LTX 2.3 Quality Is Now in Kolbo - Cinematic Video Up to 4K with Native Audio

LTX 2.3 Quality by Lightricks is live in Kolbo.AI. Text-to-video, image-to-video, and Ingredients mode for up to 9 reference images. Native audio, up to 4K, 5 to 20 seconds.

By Zohar - Kolbo.AI Team
NEW MODEL
Lightricks' high-end video tier is live. Up to 4K, native audio, and Ingredients mode to combine up to 9 reference images in one shot.

The Quality tier of LTX 2.3, Lightricks' flagship video model, is now live in Kolbo.AI. This is a meaningful step up from the standard LTX 2.3 in resolution, detail, and output length, built for creators who need finished, cinematic results straight from the generation.

Three Ways to Generate

LTX 2.3 Quality supports three distinct generation modes, and you can choose the one that fits your shot.

Text to Video. Write a prompt and generate a cinematic clip up to 20 seconds long, at resolutions up to 4K. LTX 2.3 Quality responds best to detailed, scene-specific prompts: describe the lighting, the camera angle, the mood, and what is moving. The more specific you are, the better the result.

Image to Video. Start with any still image and animate it into a video. The model reads the composition and motion cues in your image and builds the clip around them.

Ingredients. This is the mode that stands out. Give the model up to 9 reference images, each representing a different element: a character, a prop, a location, a vehicle, a texture. The model reads all of them and weaves them into a single coherent video. If you have been building Visual DNA characters or collecting reference images for a project, this is the mode that puts them to work together.

What Makes the Quality Tier Different

Higher resolution, up to 4K. The Quality tier unlocks significantly sharper output than the standard model, suitable for full-screen presentation or professional video work.

Longer clips. Generate between 5 and 20 seconds per clip, giving you enough footage to work with across a variety of content types.

Native audio. Sound is generated alongside the video, not added after the fact. The model produces ambient sound, environmental audio, and movement that fits what is happening on screen.

Cinematic prompt adherence. LTX 2.3 Quality is tuned to follow detailed prompts closely. Describe a specific shot and the model delivers it, from camera movement to scene composition to lighting quality.

What It Is Best For

  • Short films and branded content - detailed scenes with characters, props, and locations pulled from references
  • Social and commercial video - high-resolution clips with built-in audio, ready to post
  • Storyboarding and pre-visualization - fast, cinematic shots to pitch ideas before a full production
  • Product and environment video - controlled, reference-consistent shots for product campaigns or world-building

How to Try It

LTX 2.3 Quality is already in your workspace.

  1. Open the Video tool in Kolbo and choose Text to Video
  2. In the model selector, choose LTX 2.3 Quality
  3. Write a detailed cinematic prompt, or switch to Image to Video or Ingredients mode
  4. Set your resolution and clip length, then generate

LTX 2.3 Quality is live in your Kolbo workspace right now.

Generate with LTX 2.3 Quality →

Best, Zohar Founder, Kolbo.AI

Tags

videoltxlightricksnew-modeltext-to-videoimage-to-videoingredients4knative-audio

Related Posts

    We value your privacy

    We use cookies and similar technologies to improve your experience, analyze site traffic, and personalize content. You can choose which types of cookies to accept.