LTX Video 2.3 Fast
Lightricks' open-source video engine — speed-optimized, native 4K, portrait framing, and synchronized audio.
Overview
What is LTX Video 2.3 Fast
LTX Video 2.3 Fast is an open-source video generation model (text-to-video and image-to-video) developed by Lightricks, released in March 2026. It is a speed-optimized 22-billion-parameter model that generates high-fidelity video up to 4K resolution with synchronized audio and native 9:16 portrait support in a single diffusion pass.
Running it privately on Venice
On Venice, you can generate video using LTX Video 2.3 Fast without your creative prompts being stored, profiled, or used for training. Requests are routed through Venice's anonymized privacy tier, ensuring zero retention of your prompt history. This allows you to leverage its open-source power and native audio generation with complete creative sovereignty.
Assessment
Strengths and limitations
- Native audio generation: Generates synchronized ambient sound or voice in the same diffusion pass, avoiding post-processing alignment issues.
- Native portrait support: Trained on vertical data (9:16) up to 1080x1920, rather than cropping landscape video.
- High-fidelity details: Features a rebuilt VAE and a 4x larger text connector for sharper textures (hair, fabric) and tighter prompt adherence.
- Flexible durations: Supports clips up to 20 seconds long (at 1080p, 24/25 FPS).
- Open-weights sovereignty: Licensed under Apache 2.0, allowing commercial use and local deployment.
- Fast variant tradeoffs: Optimized for speed, meaning it has slightly lower motion stability and visual detail compared to the LTX 2.3 Pro variant.
- Complex physics: Still struggles with highly complex physical interactions (e.g., fluid dynamics, dense crowds) compared to larger closed-source models.
- Duration limits at high res: Clips longer than 10 seconds are restricted to 1080p at 24/25 FPS.
Samples
Sample outputs
Generated on Venice with our standard prompt suite — the same prompts we run through every model of this type, so you can judge it like-for-like.
Slow aerial drone shot gliding over a misty mountain valley at golden hour, sunlight piercing clouds onto a winding river, ancient pine forests on either side, ultra-smooth motion, professional color grading, atmospheric haze, 4K
Calm ocean waves rolling onto a black-sand beach at sunrise, golden light on wet sand, foam dissolving into the shore, a single silhouetted figure at the waterline, smooth continuous forward push, serene cinematic atmosphere
A slow tracking shot through a rain-soaked Tokyo street at night, neon reflecting in puddles, steam rising from a food cart, a person with a translucent umbrella, shallow depth of field, teal-and-orange grade, smooth steady camera
Capabilities
What it supports
- Text to video
- Image to video
- Reference to video
- Native audio generation
Variants
LTX Video 2.3 Fast model variants
LTX Video 2.3 Fast runs on Venice as 2 variants of the same underlying model. Pick by what you're starting from: a written prompt, a still image, reference images, or an existing clip. Each variant is its own model id on the API; the generation quality is the same across the family.
| Variant | What it is | Clip lengths | Resolutions | Aspect ratios | Audio | Model ID |
|---|---|---|---|---|---|---|
| Text to Videoflagship | Generate a clip from a written prompt | 6s – 20s | 1080p, 1440p, 2160p | 16:9, 9:16 | ltx-2-v2-3-fast-text-to-video | |
| Image to Video | Animate a still image into motion | 6s – 20s | 1080p, 1440p, 2160p | 16:9, 9:16 | ltx-2-v2-3-fast-image-to-video |
Capability data comes straight from the Venice model API and refreshes with every catalog ingest. The specs and pricing on this page are captured from the flagship variant; pass the model id of the variant you want to the API.
LTX Video 2.3 Fast Text to Video
Generate a clip from a written prompt. Supports clips of 6s – 20s, 1080p, 1440p, 2160p output, 16:9, 9:16 aspect ratios, with native audio.
ltx-2-v2-3-fast-text-to-videoLTX Video 2.3 Fast Image to Video
Animate a still image into motion. Supports clips of 6s – 20s, 1080p, 1440p, 2160p output, 16:9, 9:16 aspect ratios, with native audio.
ltx-2-v2-3-fast-image-to-videoSpecifications
Datasheet
- Maker
- Lightricks
- Released
- March 5, 2026
- Architecture
- Diffusion Transformer (DiT)
- Parameters
- 22 Billion
- Open weights
- Yes (Apache 2.0)
- Resolutions
- 1080p, 1440p, 2160p
- Clip lengths
- 6s – 20s in 2s steps
- Mode
- text-to-video
- Aspect ratios
- 16:9, 9:16
- Audio
- Yes
- Privacy on Venice
- Anonymized — prompts not stored
- Available on Venice since
- Mar 2026
API
Call it from your code
Venice exposes this model through the REST API. Queue a generation with the model id.
curl https://api.venice.ai/api/v1/video/queue \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "ltx-2-v2-3-fast-text-to-video",
"prompt": "Aerial drone shot over a misty mountain valley at golden hour"
}'
# Use the returned queue_id with https://api.venice.ai/api/v1/video/retrieve.
# Call /video/complete after downloading if needed.Pricing
What it costs on Venice
Pay per clip on Venice — price scales with resolution and duration (6s–20s), from $0.40.
New Venice accounts include a free daily allowance and 500 welcome credits — no credit card required.
Alternatives
How it compares
| Model | Best for | Max resolution | Native audio | Open weights | Price (Venice) |
|---|---|---|---|---|---|
| LTX Video 2.3 Fast | The open-source leader for fast, high-res video with native audio. | 4K (2160p) | Yes | Yes | from $0.40 |
| Wan 2.7 | Alibaba's open-weights model, excellent for motion but lacks native audio. | 1080p | No | Yes | from $0.55 |
| Grok Imagine | xAI's closed-source video generator, strong on prompt adherence but proprietary. | 1080p | No | No | from $0.32 |
| Seedance 2.0 | Closed-source cinematic video generator, good for storytelling. | 1080p | No | No | from $0.35 |
The open-source leader for fast, high-res video with native audio.
Use cases
What it is good for
- 01Rapid prototyping and storyboarding for video production.
- 02Creating vertical social media content (TikTok, Reels, Shorts) with native 9:16 framing.
- 03Generating short video clips with matching, synchronized background audio.
- 04High-volume, cost-effective video generation for indie creators and developers.
Prompting
Getting better results
Write prompts as a single continuous paragraph in the present tense.
Be highly specific about camera angles, lighting, and movement (e.g., 'dolly in', 'cinematic lighting').
Include atmospheric and sensory details to guide the synchronized audio generation.
Version history
Established the core architecture.
Current — speed-optimized, native portrait, 4K, and audio.
FAQ
Frequently asked questions
LTX Video 2.3 Fast is an open-source text-to-video model developed by Lightricks, released in March 2026. It is designed for high-speed, cost-effective video generation, supporting resolutions up to 4K, native 9:16 portrait aspect ratios, and synchronized audio generation in a single diffusion pass.
On Venice, pricing is pay-per-clip and scales with resolution and duration. A 6-second clip starts at $0.40 for 1080p, $0.79 for 1440p, and $1.58 for 2160p (4K). Clip lengths can range from 6 to 20 seconds.
Yes, LTX Video 2.3 Fast is an open-source model released under the permissive Apache 2.0 license, meaning it is free to self-host, fine-tune, and use for commercial projects. On Venice, you can run it via API or the web interface using Venice credits.
Yes, LTX Video 2.3 Fast features native synchronized audio generation. Unlike models that layer audio in post-processing, it processes video and audio simultaneously in a single diffusion pass, resulting in tighter alignment between sound and motion.
LTX Video 2.3 Fast is built by Lightricks and features native 4K output, native 9:16 portrait support, and synchronized audio generation. Alibaba's Wan 2.7 is another powerful open-weights model, but it typically lacks native audio generation and focuses primarily on high-quality video motion.
Venice runs LTX Video 2.3 Fast under its anonymized privacy tier. Your prompts are forwarded to a third-party provider without storing, profiling, or using them for training, ensuring your creative concepts remain private and secure.
LTX Video 2.3 Fast natively supports both 16:9 (landscape) and 9:16 (portrait) aspect ratios. Because it was trained on native vertical data rather than cropped landscape video, it delivers superior composition for mobile-first content.
The Fast variant is optimized for speed and lower cost, making it ideal for rapid prototyping, storyboarding, and quick iteration. The Pro variant offers higher visual fidelity, better motion stability, and supports advanced features like audio-to-video and video extension.
Run LTX Video 2.3 Fast privately
No prompt logging. No data used for training.