Back to Browse/Compare
SOpen

LTX-2.3

Lightricks

The open-source flagship. First open model to close the gap with proprietary leaders.

SClosed

Kling 3.0

Kuaishou

Two models in one: V3 is the prompt-driven AI Director for creative freedom. O3 is the reference-driven consistency engine for commercial production. Both share native 4K, multi-shot, and 5-language audio.

Pick LTX-2.3 if…

You want open-source audio-video, or studios needing IP control.

Pick Kling 3.0 if…

You want v3: Experimental narratives, populated group scenes (3+ chars), rapid ideation from scripts. O3: Brand advertising (product/character identity lock), serialized narratives (consistent face+voice across episodes), e-commerce (readable text+product locking). Both: multilingual commercial content, social media hooks..

Specifications

Maker
Lightricks
Kuaishou
Source Type
Open Source
Closed Source
License
Apache 2.0 (<$10M rev)
Commercial (paid tiers)
Architecture
DiT (22B) + Rebuilt VAE
Unified Multimodal (MVL) / Two models: V3 (prompt-driven) + O3 (reference-driven)
Parameters
22B
Undisclosed
Max Resolution
4K
Native 4K (3840x2160) at up to 60fps
Max Duration
20s
3-15s (up to 6 shots per generation)
FPS
Up to 50
Up to 60
Native Audio
Yes
Yes
ComfyUI Support
Yes
No
Fine-tunable
Yes
No
Min VRAM
32GB (full) / 12GB (distilled)
Cloud only
Cost / Second
$0.04
$0.10
Inputs
T2V, I2V, V2V, Audio-cond
T2V, I2V, Multi-shot (6 cuts), Element References (O3), Video Reference (O3)
On Floyo
Yes
No

Strengths & Trade-offs

LTX-2.3

Strengths

  • +22B params
  • +true 4K at 50fps
  • +first open model with synced audio
  • +rebuilt VAE
  • +native portrait

Trade-offs

  • -Full 4K needs 48GB
  • -dialogue lip-sync inconsistent
  • -in-scene text flaky

Best For

  • →Local 4K production
  • →open-source audio-video
  • →studios needing IP control

Kling 3.0

Strengths

  • +TWO MODELS: V3 = prompt-driven (3+ characters, AI Director, structured storytelling)
  • +O3 = reference-driven (Elements 3.0, video character refs 3-8s, Signature Voice binding). Native 4K at 60fps. Multi-shot up to 6 camera cuts with per-shot control (duration, size, angle, camera movement). Native lip-sync in EN/CN/JP/KR/ES + dialects + bilingual code-switching. Motion Brush for drawn motion paths. Best text rendering in AI video (signs, logos, price tags). Character identity lock across shots. Start/end frame conditioning.

Trade-offs

  • -Multi-shot not compatible with first/last frame feature. O3 optimized for 1-2 elements (V3 better for 3+ characters). Credit pricing: 12 credits/sec for 1080p+audio, 9 credits/sec for 720p+audio. Audio can be less refined than Veo. Transitions between shots can be clunky. 15s max duration.

Best For

  • →V3: Experimental narratives, populated group scenes (3+ chars), rapid ideation from scripts. O3: Brand advertising (product/character identity lock), serialized narratives (consistent face+voice across episodes), e-commerce (readable text+product locking). Both: multilingual commercial content, social media hooks.

Run these models on Floyo

Browser-based ComfyUI. No setup, no GPU required.

I2VLTX-2.3

LTX 2.3 Pro Image to Video

1.0k runs

A2VLTX-2.3

LTX 2.3 Audio to Video

124 runs

T2VLTX-2.3

LTX 2.3 Pro Text to Video

101 runs

T2VLTX-2.3

LTX 2.3 T2V (Community)

47 runs

Open Floyo →