video model · by Lightricks

LTX-2.3 AI Video Generator

Generate cinematic videos up to 4K with synchronized native audio using LTX-2.3 - Lightricks' latest DiT-based audio-video foundation model. Sharper motion, tighter dialogue lip-sync, and production-ready output, all inside Fliki.

Generated with LTX-2.3

A handful of LTX-2.3 clips generated inside Fliki. No edits, no post.

Prompt

An older Hispanic man in his late sixties of Cuban heritage, weathered hands, wearing a faded denim shirt, sits at a small kitchen table across from his teenage granddaughter, a single bare bulb hanging above casting warm pools of light, slow dolly-in framing both faces, deep amber tones, cinematic 35mm look. He pours a glass of water for her, looks up, and says quietly in Spanish, "Mija, tengo que contarte algo que nunca le he dicho a nadie." She lowers her phone slowly. SFX: water pouring, the soft tick of an old wall clock, a chair creaking, faint cicadas through an open window, a single ambient piano note underneath.

Prompt

A herd of African elephants moves slowly across a vast savannah at dusk, a mother nudging her calf along the line, dry grass swaying golden in the last light, deep cinematic 2.39:1 aerial side-tracking shot at their pace, painterly natural-history grade, soft haze, distant acacia trees silhouetted against an ember sky. A British female narrator says calmly, "For three weeks now, the matriarch has led her family toward water. Tonight, they are nearly there." SFX: low rumbling vocalisations, footfalls on dry earth, distant bird calls, a swelling solo cello and soft choral pad underneath.

Prompt

A Black woman in her late twenties, lean runner's build, wearing a fitted purple track singlet and racing flats, crouches at the starting blocks of an empty Olympic stadium at dawn, slow side-tracking shot pushing in toward her, breath visible in the cold air, deep teal-orange cinematic grade, light haze, faint stadium lights still on. She looks up at the empty stands, exhales, and whispers to herself, "Just one more time. Just for me." SFX: the sharp clack of starting blocks settling, her steady controlled breathing, a single distant bird, a low minimal pulse building underneath, no announcer.

Prompt

A vintage chrome-trimmed convertible drives slowly along a winding cliffside road on the Italian Amalfi coast at golden hour, top down, the deep cobalt Mediterranean far below, terracotta villages stacked on the cliffside, slow aerial side-tracking shot following the car, sun glinting off polished chrome, palm fronds swaying in the warm breeze, painterly travel-film grade, faint haze, cinematic 2.39:1 framing. A male voiceover says calmly in measured Italian, "Alcuni viaggi non finiscono mai davvero." SFX: the gentle hum of the engine, distant waves on the cliffs below, a soft warm jazz trumpet melody underneath, faint wind.

100M+
VIDEOS CREATED
12M+
USERS WORLDWIDE
80+
LANGUAGES SUPPORTED

Why LTX-2.3 is the latest generation

Latest LTX-2 generation

LTX-2.3 is Lightricks' newest release on the LTX-2 line - sharper detail, more stable motion, and better dialogue lip-sync than earlier versions.

Synchronized native audio

LTX-2.3 generates audio and video in a unified pass - ambient sound, effects, and dialogue arrive locked to on-screen action with natural lip-sync.

4K cinematic output

LTX-2.3 delivers up to 2160p (4K) resolution with quality suitable for final publishing, paid social, and brand campaigns - no upscaling or heavy post work needed.

Tighter motion control

Motion handling in LTX-2.3 is more stable and predictable than earlier LTX-2 builds, with fewer artifacts in high-motion scenes and cleaner subject consistency across the clip.

Open-source core, production polish

LTX-2.3 is built on Lightricks' open-source LTX-2 foundation. You get the transparency of an open model and the polish of a production-grade release.

Multi-modal inputs

LTX-2.3 supports text, image, and audio inputs for flexible creative direction. Image-to-video preserves reference quality with less drift, and audio inputs drive synchronized lip-sync.

Longer high-quality durations

LTX-2.3 handles 6, 8, and 10-second clips with consistent quality, making it suitable for narrative beats and longer branded segments.

Built for hero & campaign work

Use LTX-2.3 when the shot has to carry the campaign - paid social hero clips, product reveals, and signature brand moments. For high-volume iteration, switch to LTX-2 Fast.

How it works

How to generate a video with LTX-2.3

LTX-2.3 runs inside Fliki's standard generator. Here's the six-step flow.

Fliki prompt input showing a cinematic text-to-video description for LTX-2.3 AI video generator
Step 1

Write your prompt

LTX-2.3 rewards descriptive, directorial prompts. Mention lens feel, lighting, camera movement, subject behavior, and any dialogue you want spoken on screen.

Fliki model selector dropdown with LTX-2.3 chosen for AI video generation
Step 2

Select LTX-2.3 as your model

Pick LTX-2.3 from Fliki's model selector. Your prompt is routed to Lightricks' newest audio-video model.

Choose 16:9, 9:16, or 1:1 aspect ratio for LTX-2.3 video generation on Fliki
Step 3

Pick your aspect ratio

16:9 for landscape and cinematic web, 9:16 for vertical social. Each is composed natively in LTX-2.3.

Set video duration on the Fliki slider for LTX-2.3 multi-shot AI video generation
Step 4

Set the duration

LTX-2.3 supports 6, 8, or 10-second clips. Use the longer durations for full narrative beats, hero brand moments, and extended product reveals.

Upload an optional reference image to anchor subject and style with LTX-2.3 on Fliki
Step 5

Add a reference image (optional)

Upload a high-quality photo to guide subject, style, or composition. LTX-2.3's improved fidelity preserves reference inputs with less drift.

Pick output resolution and hit generate to create AI video with LTX-2.3 on Fliki
Step 6

Select resolution and generate

Choose 1080p, 1440p or 2160p (4K) for final delivery and hit Generate. Output is ready in Fliki with audio synced and color graded-feel baked in.

AI MODEL GALLERY

Built on the best AI models - ready inside Fliki

Every leading video, voice, and image model - integrated, unified, and tuned for creators. Generate with the latest AI video, AI voice, and AI image models from OpenAI, Google, Kling, Bytedance, ElevenLabs, and more - all from one place.

LTX-2.3 FAQ

Frequently asked questions

Everything you need to know about generating with LTX-2.3 inside Fliki.

Still curious?

Try Fliki free in your browser, no credit card required.

Start free
LTX-2.3 · Free forever plan

Generate your next hero shot with LTX-2.3.

Cinematic clips with synchronized audio, tighter motion, and 4K output. Free to start, no credit card required.

Generate your first video free

Free forever plan · No credit card required · Cancel anytime