Kuaishou (Kling AI) · Video models

Kling 3.0 Turbo

The faster variant in Kuaishou's Kling 3.0 series, for text- and image-to-video generation with no watermark by default.

Released 2026-06-17 · Last checked 2026-09-28

What’s new

Compared with standard (non-Turbo) Kling 3.0:

  • Faster output: Kuaishou positions Turbo as the faster of the two 3.0-series options, better suited to time-sensitive work.
  • Resolution instead of quality tiers: no std/pro/4K choice — you pick 720p or 1080p directly.
  • No watermark by default: the API docs state generated clips ship without a watermark by default.
  • Simpler image-to-video: just a first-frame image, with no separate aspect-ratio setting to configure — it follows the image.

What it’s good for

  • Time-boxed work like storyboard drafts or quick approval rounds.
  • Turning a single image into a short animated clip fast.
  • Higher-volume output where stable quality matters more than the top resolution tier.

Prompting tips

  • Keep prompts to around 2,500 characters (the hard cap is 3,072) — longer prompts risk truncation.
  • Duration is 3–15 seconds, default 5.
  • First-frame images have format rules: JPG/JPEG/PNG, under 50MB, short side at least 300px, aspect ratio between 1:2.5 and 2.5:1.
  • Skip the aspect ratio in image-to-video mode: it's ignored — the frame follows the source image.
  • Aspect ratio only matters for text-to-video: 16:9, 9:16, or 1:1, defaulting to 16:9.
  • For a multi-shot feel, break the prompt into a clear timed sequence — the docs describe using structured, multi-part prompts this way.
720p, 16:9, 8 seconds.
A golden retriever chases a frisbee across a lawn, slow motion catching the moment it leaps, dappled sunlight through the leaves, light and upbeat mood.

Known limits

  • Prompts max out at 3,072 characters; the docs recommend staying under about 2,500.
  • Aspect ratio has no effect in image-to-video mode — the frame is fully determined by the first-frame image.
  • First-frame images are capped at 50MB and must have a short side of at least 300px.

Using it in Nomi

On the generation canvas, add a video card, pick “Kling 3.0 Turbo” in the model picker and choose a mode (Text to video, First frame to video). Add your references, write the prompt and generate. In the storyboard you can pick it per shot. The first time, connect any one of the providers above under Settings → Models.

Sources

Pro-grade AI video. Models at their real price.

macOS · Windows · AGPL-3.0 · No account

Download Nomi
Choose the version for this computer

Nomi downloads directly when the platform is known. If the Mac chip cannot be detected reliably, choose the matching installer.

First launch on macOS

The current macOS build is not Apple Developer ID signed or notarized. Only use download links on this site or in the official Nomi GitHub repository.

  1. Download the matching DMG and drag Nomi to Applications.
  2. In Finder, right-click Nomi in Applications, choose Open, then confirm Open.
  3. If it is still blocked, open System Settings → Privacy & Security, find Nomi, and click Open Anyway.

Only if macOS says Nomi is “damaged”: confirm the installer came from an official link above, then open Terminal and run:

xattr -dr com.apple.quarantine "/Applications/Nomi.app"

You do not need to disable Gatekeeper globally, and should not do so.