Kling - ZeroGen Docs

Kling, from Kuaishou, makes video from text or frames. Each model has tiers that trade cost for quality, chosen under Tier. Pro is the default.

Model	Best for	Tiers	Sound
Kling v3	Text or frames to video, with sound	Standard, Pro, 4K	Yes
Kling v3 Turbo	Faster clips from text or a start frame	Standard, Pro	No
Kling O3	References and editing video	Standard, Pro, 4K	Yes

What you can give it

A prompt of up to 2,500 characters, or 3,072 on Kling v3 Turbo.
Kling v3: a prompt, or a start frame with an optional end frame. No references.
Kling v3 Turbo: a prompt, or a start frame (JPEG or PNG). No end frame or references.
Kling O3:
- A prompt, or start and end frames.
- Up to 4 reference images, or 7 on the 4K tier. These can be combined with frames.
- One reference video of 3 to 15 seconds, plus up to 4 images, with Video mode set to Reference media.
- Edit video: one video of 3 to 15 seconds to change, plus up to 4 images.

Output settings

Setting	Options
Aspect ratio	16:9, 9:16 or 1:1. With frames or editing, the video follows the input.
Duration	3 to 15 seconds. When editing, the video keeps the source’s length.
Audio	On or off, for Kling v3 and O3.
Quantity	1 to 4 videos.

Cost

Kling is priced per second of video, by model and tier. Sound costs more on the Standard and Pro tiers, and the 4K tier costs the most. A 5-second video on the Pro tier costs about:

Model	Estimate
Kling v3, with sound	About $1.10
Kling v3 Turbo	About $0.91
Kling O3, with sound	About $0.91

Prices are as of October 28, 2026.

​What you can give it

​Output settings

​Cost

What you can give it

Output settings

Cost