LightricksLTX 2.5
Open-Weights Video AI
LTX 2.5 is Lightricks' open-weights video model, generating text-to-video, image-to-video, and audio-to-video with native multi-shot continuity, cleaner motion, and synchronized audio up to native 4K HDR.
LTX 2.5 video generation is available on faktry: text-to-video, image-to-video, and audio-to-video in Pro and Fast tiers, with native synchronized audio and resolutions up to 2160p.
Example Videos
See LTX 2.5 in action with these example videos generated using the model.
What Is LTX 2.5?
LTX 2.5 is Lightricks' open-weights video foundation model, built as a stronger base for teams to fine-tune, self-host, or call via API. It adds native multi-shot generation that keeps character, environment, and lighting consistent across cuts, a rebuilt video decoder for cleaner motion, stronger prompt adherence, and native 4K HDR output with synchronized audio generated in the same pass.
LTX 2.5 at a glance
- Developer
- Lightricks
- Released
- 2026
- Type
- Open-weights video and audio foundation model
- Resolution
- Up to native 4K HDR
- Clip length
- Auto duration, up to 20 seconds
- Audio
- Native, synchronized generation; audio-to-video supported
- Fine-tuning
- Open weights, trainable on your own data
- On faktry
- Available now
Last updated: · Model details published by Lightricks
LTX 2.5 vs LTX 2.3
What changes between Lightricks' established LTX 2.3 and the newer LTX 2.5.
| Capability | LTX 2.3 | LTX 2.5 |
|---|---|---|
| Motion quality | Prone to motion artifacts | Rebuilt decoder, fewer artifacts, smoother motion |
| Multishot | Single-shot generations | Native connected multi-shot scenes |
| Prompt adherence | Baseline | Follows complex instructions from shorter prompts |
| Duration | Manual duration only | Auto duration matched to the requested action |
| Resolution | Up to 1080p | Native 4K HDR |
| Fine-tuning | License-limited | Dedicated pretrained checkpoint for fine-tuning |
LTX 2.5 On Faktry
Concrete LTX 2.5 operations available today, billed per second of output (or per second of input audio for audio-to-video).
Text to Video
Generate video directly from a text prompt in Pro or Fast quality.
Image to Video
Animate a starting image into motion, with an optional end frame for controlled transitions.
Audio to Video
Generate video driven by an audio track, animating an optional starting image in sync with the sound.
Fast tier resolution multipliers apply on text-to-video and image-to-video: 1x at 720p/1080p, 1.5x at 1440p, 2x at 2160p, per second.
Why Choose LTX 2.5?
Lightricks' latest open-weights video model, built for continuity, control, and production-grade motion.
Native Multishot
Generate connected shots in a single pass that hold the same character, environment, lighting, and voice across cuts.
Cleaner Motion
A rebuilt video decoder cuts motion artifacts for smoother, more natural results, even in complex scenes.
Native Synchronized Audio
Video and audio generate together in one pass, so dialogue, SFX, and ambience stay in lockstep with on-screen action.
Native 4K HDR
Render high-resolution HDR footage built for professional finishing, from fast drafts up to full 4K.
Stronger Prompt Adherence
Follows complex creative instructions from shorter, simpler prompts, with fewer retries needed to land a shot.
Open Weights, Fine-Tunable
A pretrained foundation you can fine-tune on your own data and deploy on hardware you control.
Capabilities
A closer look at what LTX 2.5 can generate and control.
Video Generation
Where LTX 2.5 Fits In
Consistent, controllable video generation for teams that need to own their pipeline.
Multi-Shot Storytelling
Generate connected scenes that hold the same character, set, and lighting across cuts instead of stitching separate clips together.
Marketing Video
Produce on-brand video content for ads and social campaigns with cleaner motion and fewer retries.
Product Visualization
Animate a product image into motion, or drive the shot from an audio track for music-synced reveals.
Social Content
Generate fast drafts or full 4K finishes from the same model, tuned to the platform and budget.
How It Works
Creating videos with LTX 2.5 is simple and powerful.
Describe Your Video
Write a text prompt, or add a starting image or audio track to drive the generation.
Configure Your Settings
Set resolution, aspect ratio, and duration, or let auto duration match the requested action.
Generate & Download
Get your video in minutes, with synchronized audio generated in the same pass.
Ready to Create with LTX 2.5?
Start generating open-weights video today with free credits.
Free credits to try us
Frequently Asked Questions
What is LTX 2.5?
LTX 2.5 is Lightricks' open-weights video foundation model, built for native multi-shot continuity, cleaner motion, stronger prompt adherence, and native 4K HDR output with synchronized audio.
What's the difference between LTX 2.5 and LTX 2.3?
LTX 2.5 adds native multi-shot generation, a rebuilt video decoder for fewer motion artifacts, automatic clip duration, and native 4K HDR, plus a dedicated checkpoint built for fine-tuning.
Can LTX 2.5 generate video from audio?
Yes. LTX 2.5's audio-to-video mode generates video driven by an audio track, optionally animating a starting image in sync with the sound, and is billed per second of input audio.
What resolutions does LTX 2.5 support?
LTX 2.5 renders from 720p up to native 4K HDR, with the Fast tier scaling cost by resolution and the Pro tier optimized for quality.
Can I fine-tune LTX 2.5 on my own data?
Yes. LTX 2.5 ships open weights with a pretrained base checkpoint built specifically for fine-tuning, so it can be adapted to your own domain and self-hosted.
Can I use LTX 2.5 videos commercially?
Yes! All videos generated with LTX 2.5 on faktry can be used for commercial purposes including marketing, advertising, and content creation.