Alibaba · video

Wan 2.6

Alibaba's open video model: up to 15 seconds at 1080p with automatic multi-shot cuts.

Text to videoImage to video
Length
5 to 15 s
Resolution
720p, 1080p
Aspect ratio
16:9, 9:16, 1:1, 4:3, 3:4
References
Up to 1 image file

Opens the Studio with Wan 2.6 and your prompt. You see the price before anything runs.

About Wan 2.6

Wan 2.6 is Alibaba's Wan video model, priced at $0.25 per second. It runs as text to video and image to video and reads prompts in Chinese and English.

Starting from text, the provider rewrites the prompt and segments it into several shots by default, so a generation can come back as a short edited sequence. Starting from an image, the image is used as the first frame. Output is 720p and 1080p and the clip runs from 5 seconds to 15 seconds. A typical generation takes about 3 min.

Good for

  • A short sequence of shots from a text prompt, cut by the model
  • Prompts written in Chinese or English
  • Animating a still that becomes the first frame

Getting good results

  1. To control the cuts, follow the provider's multi-shot format: an overall description first, then each shot in order with its time range in brackets.
  2. The aspect ratio applies when you start from text; from an image, the image is the first frame.
  3. Resolution sets the rate: $1.25 at 720p against $1.88 at 1080p for the same length.

Wan 2.6 questions

Does Wan 2.6 cut between shots on its own?

From text, yes. The provider rewrites the prompt and splits it into shots by default; you can also describe the shots yourself in the prompt.

Which lengths can I pick?

A short list from 5 seconds to 15 seconds; the price is per second, so the longest clip at 1080p costs $5.63.

Other video models

AI video generator