Video creators wanting reference-driven clips with voice, through Alibaba's API; a small free quota, but UK access is unconfirmed.

Kenning's take

What it does

Wan 2.6 is a series of Alibaba visual generation models covering text-to-video, image-to-video, reference-to-video and two image models. The new Wan2.6-R2V takes a reference clip of a person, animal or object, including its voice, and places it in new AI-generated scenes. The series adds multi-shot storytelling, better audio and video sync and clips of up to 15 seconds, and is available through Alibaba Cloud Model Studio and Wan's website.

• Wan2.6-R2V reference-to-video with appearance and voice
• Text-to-video and image-to-video upgrades
• Multi-shot storytelling with consistent visuals
• Audio-visual sync and audio-to-video
• Videos up to 15 seconds
• Two upgraded image generation models

Who it suits. Video creators and developers making short dramas, ads and other AI-generated clips.

Key facts

Category

AI models

Made by

Alibaba

Based in

China

Released

16 December 2025

How to get it

through Alibaba Cloud Model Studio (API) and the Wan website; Alibaba said it would also come to the Qwen app

Works on

API, Web

Free plan

Model Studio gives a free quota per model (50 seconds of video or 50 images, valid 90 days)

API price

wan2.6-t2v $0.10 per second at 720p, $0.15 at 1080p; wan2.6-image and wan2.6-t2i $0.03 per image (international)

FAQ

Facts from the maker's own pages and the sources listed. Checked 2026-10-11.