Generate video with Kling 3.0 Omni
The reference-first video model — up to 7 elements pinned into one shot
Kling 3.0 Omni is Kuaishou's most capable video model — its trick is the "elements" mode: you drop up to 7 reference images (a character, an outfit, a product, a background, a prop) and Kling assembles them into one coherent generated shot. On Singular Lab it runs on the same credit balance as every other video model, with the elements uploader built into the standard prompt panel — no extra tooling.
What Kling 3.0 Omni is best at
Multi-element composition: up to 7 reference images in one shot
Character continuity across ref-image variations — same person, different pose/outfit
Text-to-video AND image-to-video from the same model, one selector
Native product-shot integration (drop a product on a scene reference)
16:9, 9:16, 1:1 output at 720p / 1080p, 5-second and 10-second clips
Loud audio track included (ambient + foley baked in)
Who's using Kling 3.0 Omni
- E-commerce teams generating hero videos with the actual product on-model
- Character-driven creators keeping the same character across a mini-series
- UGC / creator ad shops where a talent + product + location need to coexist
- Storyboard-to-shot workflows: drop the storyboard frame + character sheet, get the shot
FAQ
How many reference images can I use with Kling 3.0 Omni?
+
Kling 3.0 Omni vs Seedance 2.0 — which should I use?
+
Does Kling 3.0 Omni support image-to-video?
+
What's the price per Kling 3.0 Omni video?
+
Try Kling 3.0 Omni free
Opens Manual Creation with Kling 3.0 Omni preselected.