LTX 2.3 Text to Video for Ad Film

ABOUT THE WORKFLOW

Generate an Ad Film Clip from Text
Write a shot description and get a 1080p video with synchronized audio in about 5 seconds of output. A two-pass pipeline generates at low resolution first, then upscales and refines. Toggle to image-to-video mode to animate a product photo or campaign still instead. That's it.

Model

• LTX-Video 2.3 (22B) by Lightricks. A 22B parameter DiT-based audio-video foundation model (Apache 2.0) with native audio latent support and a Gemma 3 12B text encoder for precise prompt following. Paired with the official Distilled LoRA for fewer-step generation and a 2x spatial upscaler for high-resolution output. Heavy workflow. Expect longer generation times.

HOW IT WORKS

Step 1. Write a shot description
Describe the shot like a director's brief. Include the product, camera movement, lighting, and mood. "A crystal perfume bottle sits on polished black stone surrounded by drifting water droplets. Elegant hands slowly lift the bottle while sunlight refracts through the glass creating rainbow reflections."
Works great with: product hero shots · lifestyle scenes · brand montages · mood films

Step 2. Write a negative prompt (optional)
List anything to keep out of the result. The default excludes "pc game, console game, video game, cartoon, childish, ugly." Edit it to match your brief.

Step 3. Toggle to image-to-video (optional)
Switch the mode to image-to-video (False) if you have a product photo or campaign still to animate. Upload it and the model uses it as the opening frame.

Step 4. Hit run and download
The model generates the video in two passes (low-res, then upscaled), adds synchronized audio, and returns the final 1080p result. Preview it in the workflow, then download.
Ready for: Premiere Pro · DaVinci Resolve · After Effects · any editor

First time? Leave every setting as-is. The defaults (1920×1080 · 121 frames · 24 fps) 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) — 1920×1080 · 121 frames · 24 fps · fixed seeds. About 5 seconds of 1080p video with audio. The right starting point for almost everyone.

• Want a shorter bumper or cutaway — Lower the frame length. At 24 fps, 73 frames is about 3 seconds. Good for product reveals, logo stings, and social cuts. Frame counts must be divisible by 8, plus 1.

• Want a longer hero shot — Raise the frame length. 169 frames is about 7 seconds. Useful for hero product shots and slow cinematic reveals. Longer clips take more time and credits.

• Want vertical video for social — Swap width and height to 1080×1920. The model generates native 9:16 without cropping. Ready for TikTok, Reels, and Shorts.

• Want to animate a product photo — Toggle the mode to image-to-video (False) and upload the image. The model uses your photo as the opening frame and animates from there.

• Want specific audio — Include audio direction in the prompt. "Glass clinks softly, ambient electronic music builds, a whisper says 'Discover'" gives the model specific audio cues.

• Want a different take — Change both Seed Pass 1 and Seed Pass 2 together. Each seed pair produces a different interpretation of the same brief.

Prompt: Write like a director's shot list. Include the product, surface, lighting, camera move, and audio. "Extreme macro of a watch dial on brushed steel, the second hand ticks, camera pulls back slowly revealing the full case, warm tungsten sidelight, soft mechanical ticking sound" works better than "luxury watch video." The more specific the brief, the closer the output lands to what you need.

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

💎 Product Hero Shots
Generate cinematic product reveals from a text brief. Perfume bottles refracting light, sneakers rotating on concrete, skincare gliding across water. No studio, no photographer, no rig.

📱 Social Ad Cuts
Produce vertical 9:16 clips for paid social. Write the brief, generate the clip, export to Premiere or directly to the ad platform. Fast iteration means more variants per campaign.

🎬 Pitch Films and Mood Reels
Generate concept footage to sell a creative direction before committing to a full production budget. Show a client the look, camera language, and pacing of a campaign idea in minutes.

🛍️ E-commerce and Lifestyle Video
Turn product descriptions into lifestyle video clips with ambient sound. A candle flickering on a side table, a handbag placed on a cafe counter. Ready for product pages and email campaigns.

🎧 Audio-Visual Brand Content
Generate brand films with matched audio in one pass. Ambient sound, music cues, and speech-like audio are produced from the prompt. No separate sound design step for early-stage content.

WHAT WORKS BEST / WHAT TO AVOID

✅ Works great

• Detailed shot descriptions with product, surface, lighting, and camera direction

• Slow, deliberate camera moves (dolly, orbit, push-in, rack focus)

• Single-product hero shots on clean surfaces

• Prompts with specific audio cues

⚠️ May produce softer results

• Vague briefs like "luxury product ad"

• Rapid cuts or multi-scene sequences in a single clip

• Very long clips (consistency degrades with length)

• Specific brand logos or legible product text (the model generates approximations, not exact typography)

FAQ

What is LTX-Video 2.3?
LTX-Video 2.3 is a 22B parameter open-source audio-video foundation model by Lightricks, released under the Apache 2.0 license. It generates synchronized video and audio in a single pass using an asymmetric dual-stream Diffusion Transformer with a 14B video stream and a 5B audio stream. This workflow is configured for ad film production with text-to-video as the default mode.

Can I generate product videos without a product photo?
Yes. The workflow defaults to text-to-video mode. Describe the product, the surface, the lighting, and the camera move, and the model generates the entire shot from text. Toggle to image-to-video mode when you have a specific product photo to animate.

Does the model generate audio for ad clips?
Yes. LTX 2.3 generates synchronized audio alongside the video. Include audio direction in your prompt: "glass clinks, ambient electronic music, a soft whisper." The model produces environmental sounds, music cues, and speech-like audio matched to the visual scene. For final delivery, you will likely replace this with licensed music, but the generated audio is useful for pitch films and internal reviews.

What does two-pass upscaling mean?
The workflow generates the video at low resolution first (768×512), then runs a second pass with the 2x spatial upscaler to refine and sharpen the output to 1080p. The two-pass approach produces sharper detail in product surfaces, reflections, and textures than generating at full resolution in a single pass.

Can I use the output in client deliverables?
Yes. LTX-Video 2.3 is released under the Apache 2.0 license, which allows commercial use, modification, and redistribution. You can use the outputs in client pitches, paid campaigns, broadcast, and published content.

How do I get consistent shots across a campaign?
Lock both seeds to reproduce the same shot. Change the prompt to adjust framing, angle, or product placement while keeping the same visual treatment. For multi-shot consistency, describe the same lighting setup and surface across prompts and keep the seeds fixed.

How to run LTX 2.3 for ad film online?
You can run LTX 2.3 for ad film online through Floyo. No installation, no setup, no API key to wire up. Open the workflow in your browser, write a shot description, 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 creative director generates a product shot and likes the result. A producer opens that exact run from shared history and sends it to the client. 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 shot description and run it. The settings are already set.

→ Launch Workflow, Free

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