LTX-2.3 Pro · Image to Video With Audio

ABOUT THE WORKFLOW

Animate a Still With Sound at Up to 4K
Upload a start image, add an end image if you want, and describe the shot. LTX-2.3 Pro builds the picture and the audio together in one pass and returns a clip with sound. Duration, resolution, aspect ratio, and frame rate are all adjustable from the node.

Partner node. This workflow calls an external API, so each run uses credits from your API wallet. No API key needed. Floyo handles the connection.

Model

• LTX-2.3 Pro by Lightricks. The quality tier of the LTX-2.3 family, a 22 billion parameter model with a Gemma 3 12B text encoder. Pro adds higher fidelity, smoother motion, and native 4K output over the open-weight checkpoints. Video and audio are written in the same pass.

HOW IT WORKS

Step 1. Upload your start image
The frame the clip opens on. It sets the subject, the scene, and the framing.
Works great with: character art · product shots · landscapes · concept frames

Step 2. Upload an end image (optional)
The frame the clip should land on. Both images should frame the subject at a similar size. Leave it empty and the model picks where the shot ends.

Step 3. Write the shot
Describe the action, the camera, and the sound. Long, detailed prompts suit this model. Write the visual action first, then the soundscape, then the score.

Step 4. Set resolution and length
Pick from 720p to 2160p and up to 10 seconds. Resolution and duration together drive what a run costs.

Step 5. Hit run and download
The clip comes back as an MP4 with sound, saved under Ltx.
Ready for: Premiere · DaVinci Resolve · CapCut · After Effects

First time? Leave every setting as-is. The defaults (2160p · 6 seconds · auto aspect ratio · 25 fps · audio on) are the right starting point for almost everyone.

RECOMMENDED SETTINGS

Quick-start guide. Find the goal that matches yours and copy the settings.

• Standard clip (most people) — 2160p · 6 seconds · auto aspect ratio · 25 fps · audio on. The right starting point for almost everyone.

• Quick test before a 4K run — Drop to 720p at 6 seconds. Cheapest way to check whether the motion and the sound read the way you wanted before committing to the full resolution.

• Need a longer clip — Raise duration up to 10 seconds. Duration and resolution together drive cost, so raise one at a time.

• Need a specific shape — Switch aspect ratio from auto to the shape you want. Auto follows your start image.

• Want the clip to land on a specific frame — Upload a second image as the end frame. The model interpolates between the two. Frame the subject at a similar size in both images.

• Want smooth slow motion — Set FPS to 50. The model generates at the frame rate you pick, so 50 gives you true high-frame-rate footage rather than duplicated frames.

• Sound is not matching the action — Describe the sound separately in the prompt. Name the room tone, the effects, and the score as their own lines after the visual action.

• Playback is slightly fast — The save node runs at 27 fps while the model outputs at 25. Set the save step's frame rate to match your chosen FPS to fix the speed.

Prompt: Write it as a timeline. Visual action first: "camera slowly pushes in from a side profile, wind blows through his hair and coat, light rain and floating embers drift." Then sound: "rolling thunder, rain on leaves, distant wind." Then score if you want one: "deep orchestral strings, building tension." Splitting the brief this way gives the model a clear read on what to show and what to play.

LEARN

📹 Videos

• Intro to Floyo

• ComfyUI 101 Free Course ft. Sebastian Kamph

• Floyo 101 for Team Collaboration

✨ Quick links

• Help docs

• LoRA Training guide

• Discover more workflows

USE CASES

🎬 Cinematic Spots
Turn a concept frame into a 4K clip with camera moves, sound effects, and a score already on the track.

📢 Product and Brand Films
Animate a product still into a finished 4K spot without a shoot, a sound session, or a separate audio pass.

🎞️ Start-to-End Interpolation
Give the model a first frame and a last frame and get a smooth transition between the two with matching sound.

🎵 Sound-Led Scenes
Write the soundscape and the score as part of the prompt when the audio matters as much as the picture.

WHAT WORKS BEST / WHAT TO AVOID

✅ Works great

• Clear start images with a readable subject and depth

• Prompts that name the camera move and describe the sound separately

• Start and end images framed at a similar subject size

• Clips of 6 to 10 seconds at 720p to 2160p

⚠️ May produce softer results

• Prompts that describe the frame rather than what happens

• Start and end images with wildly different framing or subject sizes

• Complex dialogue with precise lip sync

• Leaving the sound description out and hoping for audio

FAQ

What is LTX-2.3 Pro?
LTX-2.3 Pro is the quality tier of Lightricks' LTX-2.3 model family, released 5 March 2026. It is a 22 billion parameter diffusion transformer with a Gemma 3 12B text encoder that generates synchronized video and audio in one pass. Pro adds higher fidelity, smoother motion, and native 4K at up to 50 fps over the open-weight distilled and dev checkpoints.

How is LTX-2.3 Pro different from the open-weight LTX-2.3 workflow?
The open-weight workflow runs the distilled checkpoint locally on your own GPU through a multi-stage render and upscale pipeline. Pro calls the hosted Lightricks endpoint, which runs the full-fidelity Pro checkpoint at native resolution without a local upscale chain. Pro reaches 4K natively and handles up to 10 seconds in a single call. Use the open-weight workflow for local control and no per-run API cost, and Pro when you want the highest fidelity at 4K without managing the pipeline yourself.

Does LTX-2.3 Pro generate audio with the video?
Yes. Video and audio are generated in the same pass, so dialogue, effects, and room tone land on the same timeline as the picture. Describe the sound in your prompt and it goes onto the track. Turn audio off if you want a silent clip.

What resolution and length does LTX-2.3 Pro support?
720p, 1080p, and 2160p at 24, 25, or 50 frames per second, up to 10 seconds per run. Resolution and duration together drive what a run costs, so a 4K 10-second clip is the most you can spend in one call.

Can LTX-2.3 Pro interpolate between a start and end frame?
Yes. Upload a second image as the end frame and the model builds smooth motion from the first to the last. Both images should frame the subject at a similar size for the cleanest result.

Is LTX-2.3 Pro free for commercial use?
The open weights are released under the LTX-2 Community License, which is free for organizations under 10 million USD in annual revenue. This workflow calls the hosted Pro endpoint rather than running the open weights, so the endpoint's own terms apply. Check the current terms before building on it.

How to run LTX-2.3 Pro online?
You can run LTX-2.3 Pro online through Floyo. No installation, no setup, no API key to wire up. Open the workflow in your browser, upload your image, write the shot, and hit run. Free to try.

WHY FLOYO?

Floyo is the only platform with team collaboration for ComfyUI in the browser. You run workflows with no install. You share run history, assets, and models across your team. You pay only when you generate. Floyo supports open-source and closed-source models.

A designer runs an edit and likes the result. A teammate opens that exact run from shared history and keeps going. No file handoffs. No version confusion.

For studios and enterprise teams, Floyo adds private workspaces, pooled resources, and a team usage dashboard. Other ComfyUI cloud tools run for one person at a time. Floyo runs for the whole team, with transparent per-generation costs.

Ready to try it?
Upload a start image, write the shot and the sound, and run it. The settings are already set.

→ Launch Workflow, Free

Questions? Watch the free course or check the FAQ above.