Video generation
Available now

LTX 2.5

by Lightricks

Lightricks' video model with sound built in.
Clips, audio‑led scenes and talking portraits.

LTX 2.5

Key features

Technical specifications

720p to 4K

720p or 1080p on LTX 2.5, up to 2160p on LTX 2.5 Fast

6 to 20s

6, 8 or 10 seconds on LTX 2.5, up to 20 seconds on Fast

Native

Made with the picture, switchable on text and image to video

4 modes

Text, image and audio to video, plus lip sync

Use cases

Short ads with sound

A product moment with its own pour, click or room tone, ready to post without a separate sound pass.

Audio-led visuals

Feed a short beat, sting or voice line and the motion follows it, for teasers, loops and title moments.

Talking portraits

One photo and a recorded line become a speaking character for explainers, intros and versions in other languages.

Vertical social clips

Native 9:16 output framed for the phone from the first frame, ready for Reels, Shorts and other vertical feeds.

Longer takes on Fast

Up to 20 seconds in one generation on LTX 2.5 Fast, for establishing shots and slow reveals that need room.

4K for big screens

Render at up to 2160p on LTX 2.5 Fast when a shot has to hold up on a large display or survive a crop.

Prompt examples

Rain on a roof

Rain on a roof

Close-up of rain hitting a tin roof at dusk, steady drumming, distant thunder

Edit prompt
Product pour

Product pour

Slow pour of cold brew over ice in a glass, crackle of ice, soft cafe murmur

Edit prompt
Night alley

Night alley

Handheld walk down a wet neon-lit alley at night, footsteps and far traffic

Edit prompt
Coastline

Coastline

Aerial drift over a rocky coast at sunrise, waves booming against the cliffs

Edit prompt
Workshop

Workshop

A potter's hands shaping wet clay on a spinning wheel, soft hum and splash

Edit prompt
Skate session

Skate session

Low tracking shot of a skateboarder carving an empty pool, wheels rumbling

Edit prompt

Overview

LTX 2.5 is the newest video model in Lightricks' open LTX family, released in August 2026 with weights on Hugging Face. It generates video and audio together in one pass, so dialogue, effects and ambience arrive already matched to the picture. It replaces LTX 2.3 on Morphic.

What LTX 2.5 does differently

Most video models treat sound as an afterthought or skip it. LTX 2.5 starts from the idea that a clip should leave the generator finished, picture and sound together, and it also runs the other way round: give it a short audio clip and it builds the video around that track. The same family covers lip sync, so a single photo and a recorded line become a speaking character.

LTX 2.5 and LTX 2.5 Fast on Morphic

Morphic offers two tiers. LTX 2.5 renders 6 to 10 second clips at 720p or 1080p. LTX 2.5 Fast reaches 2160p and runs up to 20 seconds, for longer establishing shots and frames that need to hold up on a big screen. Both work from text or a first-frame image in 16:9 or 9:16, with native audio you can switch off. Audio to video and lip sync use LTX 2.5 and are on the Basic plan and above. Pick either tier in Copilot or on the Canvas, next to models like Veo, Kling and Seedance, and compare takes side by side.

Simple pricing

Get started for free today, with the option to upgrade or cancel anytime.

Basic

$19/ month
billed as $228 per year

2400 monthly credits

1 user only

All models

Workflows

Standard

$24/ month
billed as $288 per year

3625 monthly credits

1 user only

All models

Workflows

Pro

$45/ month
billed as $540 per year

6350 shared monthly credits

1 user

+ up to 4 more at extra cost

All models

Workflows

Max

$170/ month
billed as $2040 per year

24650 shared monthly credits

1 user

+ up to 9 more at extra cost

All models

Workflows

Enterprise

For higher limits

Custom

pricing and billing terms

High-volume credits
Custom seat limits
All models
Workflows
Pricing Gradient

Free

For playing around

$0

forever free

Up to 20 credits
1 user only
Limited models
Workflows
Compare plan details

FAQs

What is LTX 2.5?
LTX 2.5 is Lightricks' video generation model, released in August 2026 with open weights on Hugging Face. It generates the picture and the sound together. On Morphic it covers text to video, image to video, video built around an audio clip, and lip sync from a photo.
What is the difference between LTX 2.5 and LTX 2.5 Fast?
Both make video with native audio from text or an image. LTX 2.5 renders at 720p or 1080p in clips of 6, 8 or 10 seconds. LTX 2.5 Fast goes further on size and length: 720p up to 2160p (4K), and 6 to 20 seconds in two-second steps.
Does LTX 2.5 generate audio?
Yes. Dialogue, sound effects and ambience are generated in the same pass as the picture, so they stay tied to what happens on screen. On text and image to video you can switch audio off and get a silent clip instead.
How does audio to video work with LTX 2.5?
Attach one audio clip between 2 and 10 seconds long, such as a beat, a sting or a spoken line, and optionally an image to start from. LTX 2.5 builds the video around that audio and keeps it as the soundtrack. Audio to video is on the Basic plan and above.
Can LTX 2.5 do lip sync?
Yes. Give it one image of a face and an audio clip of 2 to 10 seconds, and it animates the face to speak or sing the audio, with a prompt to steer the performance. Lip sync with LTX 2.5 is on the Basic plan and above.
What resolutions, lengths and aspect ratios does LTX 2.5 support?
LTX 2.5 renders 720p or 1080p (default 1080p) at 6, 8 or 10 seconds. LTX 2.5 Fast renders 720p, 1080p, 1440p or 2160p at 6 to 20 seconds. Both make 16:9 or 9:16 video, and image to video can also follow the shape of your image.
What happened to LTX 2.3 on Morphic?
LTX 2.5 replaced LTX 2.3 on Morphic. A saved workflow, preference or prompt that names LTX 2.3 now runs on LTX 2.5, and the reply tells you which model ran.
Which Morphic plans include LTX 2.5?
Text to video and image to video with LTX 2.5 and LTX 2.5 Fast are on every plan, Free included. Audio to video and lip sync need the Basic plan or above. See the pricing page for current plans.
How do I use LTX 2.5 on Morphic?
Open Copilot or the Canvas and pick LTX 2.5 or LTX 2.5 Fast in the video model list. Type a prompt, or attach an image for image to video, or an audio clip for audio to video. Because it runs beside Veo, Kling and Seedance, you can send one prompt to several models and keep the best take.

Other models

Gemini Omni Flash 1.2

Google DeepMind

Google's video model, now fifteen seconds long. Edit the whole take for 30% less a second.

ChatGPT Images 2.6

OpenAI

OpenAI's image model, released November 2026. Animated GIFs, SVG export, faster drafts.

SAM 3.1

Meta

Meta's Segment Anything model. Track any subject as a mask that follows it.

Seed Audio 1.5

ByteDance

ByteDance's full-scene audio model. Dialogue, music and effects, as separate tracks.