LTX-2.5 Fast · Text to Video

ABOUT THE WORKFLOW

Generate a Video with Sound from Text
Describe a scene and get a video with synchronized audio back. LTX-2.5 Fast is the speed-optimized tier of the LTX-2.5 family, rendering faster than real time on most configurations. Supports up to 4K resolution and clips up to 20 seconds.

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-2.5 Fast by LTX (Lightricks). A 22B-parameter open-weights video model released August 2026. Fast is the speed tier, generating video and audio in one pass at up to 4K. Trades some fidelity versus Pro for faster rendering and higher resolution output.

HOW IT WORKS

Step 1. Write your prompt
Describe the scene you want. Subject, action, camera move, lighting, mood, and the sound you want to hear. The prompt field is empty by default.
Works great with: cinematic scenes · product clips · social content · motion design

Step 2. Pick duration, resolution, and ratio
Set clip length (up to 20 seconds), resolution (up to 4K), and aspect ratio. Defaults are 6 seconds, 1080p, 16:9.

Step 3. Set camera motion (optional)
Choose a camera movement preset from the dropdown, or leave it on none and describe the camera in your prompt instead.

Step 4. Hit run and download
The model generates the video with synced audio and returns it as a video file. Preview it in the workflow, then download.
Ready for: Premiere · DaVinci Resolve · After Effects · any NLE

First time? Leave every setting as-is. The defaults (6 seconds · 1080p · 16:9 · 25 fps · audio on · no camera motion) are the right starting point for almost everyone.

RECOMMENDED SETTINGS

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

• Standard generation (most people) — 6 seconds · 1080p · 16:9 · 25 fps · audio on. The right starting point for almost everyone.

• Want a vertical video — Switch aspect ratio to 9:16 for stories, reels, or portrait content.

• Want a longer clip — Raise duration up to 20 seconds. Fast supports longer clips than Pro. Duration drives cost.

• Want 4K output — Raise resolution to 4K. Combine with shorter duration to keep costs down. 4K at 20 seconds is the most expensive configuration.

• Want a specific camera move — Pick a preset from the camera motion dropdown, or describe the move in your prompt.

• Want to iterate fast before finishing on Pro — Use Fast at 1080p and short duration to test prompts. Once you find the right brief, switch to the Pro workflow for higher fidelity on the final take.

• The video does not match the prompt — Rewrite the prompt before changing settings. Name the subject, the action, the camera, and the lighting. Include sound cues.

Prompt: Write it like a shot brief. "A woman walks through a rainy Tokyo street at night, neon reflections on wet pavement, slow tracking shot, city ambience" works better than "woman in the rain." Include what you want to hear. The model generates audio alongside the picture, so sound cues shape the result.

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

🎬 Rapid Previsualization
Test scene direction, camera angles, and pacing before committing to a full production shoot or a slower, higher-fidelity render on Pro.

📱 Short-Form Social Content
Generate clips with sound for reels, stories, or ads. Vertical 9:16 is one dropdown away. Audio is included, so the clip is ready to post.

🎥 4K Footage Drafts
Generate at 4K when the final output will be viewed on a large screen or projected. Fast is the only LTX-2.5 tier that supports 4K text-to-video.

🔊 Audio-Visual Prototyping
The model generates stereo audio in the same pass as the video. Describe both the picture and the sound to get a clip that carries its own atmosphere without a separate audio step.

WHAT WORKS BEST / WHAT TO AVOID

✅ Works great

• Shot-brief style prompts with subject, action, camera, and sound

• Short clips for fast iteration (5 to 6 seconds)

• 4K output for large-screen or projection use

• Prompts that describe what the camera and subject do

⚠️ May produce softer results

• Vague one-line prompts with no camera or action direction

• Complex scenes with crowds or fast motion (Pro handles these better)

• Maxing out duration, resolution, and 4K at the same time (raise one at a time)

• Expecting Pro-level fidelity on fine detail (Fast trades some sharpness for speed)

FAQ

What is LTX-2.5 Fast?
LTX-2.5 Fast is the speed-optimized tier of Lightricks' LTX-2.5 video model, released August 2026. It is a 22B-parameter open-weights model that generates video with synchronized audio from a text prompt, at up to 4K resolution and 20 seconds per clip. It renders faster than real time on most configurations.

What is the difference between LTX-2.5 Fast and LTX-2.5 Pro?
Fast is the speed tier. It supports up to 4K resolution and clips up to 20 seconds, and renders faster. Pro is the quality tier, capping at 1080p with higher fidelity per frame. Use Fast for rapid iteration, higher resolution, or longer clips. Use Pro when visual detail on faces, text, or complex scenes matters most.

Does LTX-2.5 Fast generate audio?
Yes. The model generates synchronized stereo audio alongside the video in one pass. This workflow carries the audio through to the output, so the downloaded video includes sound. Include sound cues in your prompt to shape what you hear.

What resolution and duration does LTX-2.5 Fast support?
Up to 4K resolution and clips up to 20 seconds at 25 fps. Higher resolution and longer duration both increase cost per run. The default configuration is 1080p at 6 seconds.

Is LTX-2.5 Fast open source?
LTX-2.5 is released under the LTX-2.x Community License, which allows free use for individuals and organizations under $10M annual revenue. Above that threshold, a commercial license is required. Review the license terms on Hugging Face before using outputs in production.

Can LTX-2.5 Fast generate multishot scenes?
Yes. Native multishot is a core feature of LTX-2.5, meaning the model can generate connected scenes with different camera angles in a single generation while holding character and environment consistent across cuts. Describe the shots in your prompt.

How to run LTX-2.5 Fast online?
You can run LTX-2.5 Fast online through Floyo. No installation, no setup, no API key to wire up. Open the workflow in your browser, write a prompt, 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 director runs a take 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?
Write a scene brief and run it. Video with audio, back in seconds.

→ Launch Workflow, Free

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