Text to Video AI Generator

Describe a scene in words and generate a video clip with leading AI video models, with sound on supported models.

First frame
(optional) Switch models to add images

0 / 2000

Sound

Video settings

Aspect ratio

Duration

Resolution

Image input

Model

Seedance ByteDance

Kling Kuaishou

How to write a text to video prompt

State who or what is on screen and what happens: a runner crossing a bridge, steam rising from a cup.

Add the setting and the camera movement: close-up, slow push-in, aerial shot, handheld.

Finish with lighting and tone: golden hour, neon night, soft and calm, fast and energetic.

Compare what each text to video model available on LubanAI can generate.

Model	Sound	Duration	Resolution
MiniMax H3 Max	Always on	5–15s	480p / 768p / 1080p
MiniMax H3	Always on	4–15s	768p / 2K
Seedance 2.5	Always on	4–15s	480p / 720p / 1080p
Seedance 2.0 Mini	Always on	4–15s	480p / 720p
Seedance 2.0 Fast	Always on	4–15s	480p / 720p
Seedance 2.0	Always on	4–15s	480p / 720p / 1080p
Wan 3.0	On / off	2–15s	480p / 720p / 1080p
Wan 3.0 Prime	On / off	2–15s	480p / 720p / 1080p
Kling 4.0	On / off	3–30s	720p / 1080p
Kling 3.0	On / off	3–15s	720p / 1080p / 4K
Kling 3.0 Turbo	Always on	3–15s	720p / 1080p
Veo 3.1 Fast	Always on	4–8s	720p / 1080p / 4K
Veo 3.1	Always on	4–8s	720p / 1080p / 4K

Models marked in the sound column generate audio together with the video. Where sound is optional, turn it on or off before generating.

Clip length depends on the model. The table above lists the duration range for each one.

Yes. Every video model on LubanAI offers the vertical 9:16 aspect ratio.

Yes. Add a first frame in the same editor, or see the image to video page for the models that accept a photo.

Yes. If generation fails, the credits for that video are refunded automatically.