Wan 2.7
Wan 2.7 is a Tongyi Wanxiang generative model family that helps creators and design teams turn text, images, or references into videos and edited images.
Tool overview
Adoption judgment: Wan 2.7 is worth testing, but the available evidence proves attention more strongly than dependable quality. Most of the 10 sources are short X posts, reposts, platform promotion, or leaderboard discussion. They are heat signals, not proof of usability. The stronger capability evidence consists mainly of two longer Zhihu articles with some hands-on observations, so Wan 2.7 is a promising candidate rather than a fully validated production standard.
The family is described as supporting text-to-video, image-to-video, reference-to-video, image generation, and image editing. Social posts show examples involving comic storytelling, ads, dance, and short-form video. The Zhihu articles focus on Wan2.7-Image’s face diversity, palette control, interactive editing, and text rendering. One article reports a strong human-preference blind-test result; another says its control is a noticeable strength while pure visual aesthetics and extreme photorealism still have room to improve. These are reported observations, not universal guarantees.