VideoAnonymized

LTX Video 2.3 Full Quality

LTX Video 2.3 Full Quality is an open-weights, audio-visual foundation model delivering high-fidelity 4K video with synchronized sound, native portrait output, and strong prompt adherence — now on Venice with zero retention.

Generate videoGet API key

What is LTX Video 2.3 Full Quality?

LTX Video 2.3 Full Quality is Lightricks' open-weights, diffusion-based AI model that generates synchronized video and audio from text, images, or audio input. Released March 5, 2026, it supports up to 20-second clips in 4K, native portrait aspect ratios, and HDR output via LoRA, with improved fine detail and motion coherence over its predecessor.

Use LTX Video 2.3 Full Quality privately on Venice

On Venice, LTX Video 2.3 Full Quality runs under an anonymized privacy tier — your prompts are not stored, profiled, or used for training. This means you can generate commercial-grade video content with full sovereignty, knowing your creative concepts remain private. The model's open weights and uncensored nature allow for permissionless innovation, whether you're building branded content or experimental media.

Anonymized
No prompt training
TEE · hardware enclave
End-to-end encrypted

What can LTX Video 2.3 Full Quality do?

Strengths
  • Open weights enable local execution, fine-tuning, and full ownership of output — rare among high-fidelity video models.
  • Native portrait generation (up to 1080×1920) trained on vertical data, not cropped from landscape — ideal for social media workflows.
  • Synchronized audio generation where voice, music, and sound effects shape motion and pacing — not added in post.
  • Improved spatial intelligence and motion coherence, with sharper fine detail thanks to a rebuilt VAE and latent space.
  • Strong prompt adherence with a 4x larger text connector, handling complex prompts involving multiple subjects and spatial relationships.
Limitations
  • Complex physics (e.g., water, crowds) and emotional tonal subtlety still lag behind top closed models like Kling 3.0 or Runway Gen-4.5.
  • Image-to-video mode can exhibit instability in current release, requiring careful prompt engineering or retakes.
  • Limited to 10-second clips at 1440p/4K in Pro tier; 20-second generation only available at 1080p.
  • No end-to-end encryption or TEE protection on Venice — privacy is anonymized, not fully secured.

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.

Cinematic landscape

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

Seamless loop

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

Urban cinematic

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

Compare every video model on these prompts

LTX Video 2.3 Full Quality model variants

LTX Video 2.3 Full Quality 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.

VariantWhat it isClip lengthsResolutionsAspect ratiosAudioModel ID
Text to VideoflagshipGenerate a clip from a written prompt6s, 8s, 10s1080p, 1440p, 2160p16:9, 9:16ltx-2-v2-3-full-text-to-video
Image to VideoAnimate a still image into motion6s, 8s, 10s1080p, 1440p, 2160p16:9, 9:16ltx-2-v2-3-full-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 Full Quality Text to Video

Generate a clip from a written prompt. Supports clips of 6s, 8s, 10s, 1080p, 1440p, 2160p output, 16:9, 9:16 aspect ratios, with native audio.

ltx-2-v2-3-full-text-to-video

LTX Video 2.3 Full Quality Image to Video

Animate a still image into motion. Supports clips of 6s, 8s, 10s, 1080p, 1440p, 2160p output, 16:9, 9:16 aspect ratios, with native audio.

ltx-2-v2-3-full-image-to-video

How to use LTX Video 2.3 Full Quality via API

Venice exposes this model through the REST API. Queue a generation with ltx-2-v2-3-full-text-to-video.

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-full-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.

Specifications

MakerLightricks
ReleasedMarch 5, 2026
ArchitectureDiT-based diffusion model
Parameters22 billion
Resolutions1080p, 1440p, 2160p
Clip lengths6s, 8s, 10s
Modetext-to-video
Aspect ratios16:9, 9:16
AudioYes
Privacy on VeniceAnonymized — prompts not stored
Available on Venice sinceMar 2026
LicenseLTX-2 Community License

Pricing

Pay per clip on Venice — price scales with resolution and duration (6s–10s), from $0.53.

1080p · 6s
$0.53
1440p · 6s
$1.06
2160p · 6s
$2.11

New Venice accounts include a free daily allowance and 500 welcome credits — no credit card required.

LTX Video 2.3 Full Quality vs alternatives

ModelMax resolutionStrongest atOpen weightsPrice (Venice)
LTX Video 2.3 Full Quality2160pOpen, audio-visual sync, portraitYesfrom $0.53
Wan 2.7 EnhancedCinematic realismYesfrom $0.68
Kling O3 ProHigh-motion scenesNofrom $0.46
Vidu Q3Long-duration clipsNofrom $0.27

The only open, high-fidelity audio-video model on Venice with native portrait support.

What is LTX Video 2.3 Full Quality good for?

  • Social-first video content with native vertical framing and synchronized audio.
  • Podcasts, avatars, and voice-driven clips where audio defines visual pacing.
  • Branded content and marketing videos requiring high fidelity and prompt control.
  • Local, on-prem video generation for enterprises concerned with IP leakage or cloud costs.
  • Creative experimentation with open, modifiable video generation models.

Prompting tips

  • Use clear spatial and stylistic descriptors — 'a woman in red dress on left, city skyline on right' — to leverage improved prompt adherence.
  • For portrait content, specify 'vertical video' or '9:16 aspect ratio' to ensure native generation.
  • Use HDR LoRA for grading-ready EXR output if post-production flexibility is needed.
  • For audio-to-video, provide clean audio input with clear pacing cues to maximize visual synchronization.

Version history

LTX-2
2026-01

Predecessor model with foundational architecture.

LTX Video 2.3 Full Quality
2026-03

CurrentCurrent — improved VAE, audio sync, native portrait, HDR.

Frequently asked questions

LTX Video 2.3 Full Quality is Lightricks' open-weights AI video model released on March 5, 2026. It generates up to 20-second clips in 4K with synchronized audio from text, images, or audio input, featuring improved detail, motion, and prompt adherence over LTX-2.

On Venice, pricing starts at $0.53 per clip for 1080p at 6 seconds, scaling with resolution and duration. Higher resolutions like 2160p cost up to $2.11 for a 6-second clip.

It is open weights under the LTX-2 Community License, allowing commercial use, modification, and local execution. While not fully open source in the traditional sense, it is freely available and Apache 2.0-compatible for derivative works.

Yes, it natively generates synchronized audio — including voice, music, and sound effects — as part of the video generation process, not added in post.

It supports 1080p, 1440p, and 2160p (4K) resolutions, with aspect ratios of 16:9 and 9:16. Native portrait generation is trained on vertical data, not cropped.

Yes, the variant 'ltx-2-v2-3-full-image-to-video' animates still images into motion with up to 10 seconds of video at 1080p, 1440p, or 2160p, including synchronized audio.

LTX excels in audio-visual sync, open weights, and native portrait output, while Wan 2.7 Enhanced leads in cinematic realism and motion stability. Choose LTX for open, audio-driven workflows; Wan for pure visual fidelity.

Yes, via an IC-LoRA, LTX Video 2.3 Full Quality can generate directly in HDR or convert SDR footage to EXR, providing greater grading latitude for professional finishing pipelines.

Up to 20 seconds in 1080p. At higher resolutions (1440p and 2160p), maximum clip length is 10 seconds.

Related models

Run LTX Video 2.3 Full Quality privately.

No prompt logging. No data used for training. Free to start — no credit card.

Room