Wan 2.7 Uncensored

• 🎬 Alibaba's Wan 2.7, image-to-video, in an uncensored variant
• 🆕 Supports first-and-last-frame interpolation control
• 👁️ Animates a start image with motion prompts
• 🔧 Supports first-frame, first-and-last-frame, and video continuation
• 🎯 Strong subject coherence across frames per catalog
• 🏢 Built by Alibaba, released 2026

Alibaba

Alibaba Group is a Chinese multinational technology company founded in 1999 and headquartered in Hangzhou, Zhejiang. Originally built around e-commerce and cloud computing, Alibaba has become one of the most prolific contributors to open-weight AI research,…

Wan 2.7 Uncensored is an image-to-video model in Alibaba's Wan family, here offered as a content-unrestricted variant. It takes a still image as the first frame, plus a text prompt, and animates it into a short clip. According to Alibaba Cloud's documentation, the Wan image-to-video line supports first-frame-to-video, first-and-last-frame interpolation, and video continuation.

Compared with its same-family predecessors Wan 2.6 and Wan 2.5 Preview, this release belongs to the 2.7 generation of the line. Among the documented controls is first-and-last-frame interpolation, where you supply both a start and end image and the model builds a transition between them.

Venice's catalog summarizes the family as offering photorealistic generation with strong subject coherence across frames. The model pairs with the Wan 2.7 Uncensored text-to-video sibling and the standard Wan 2.7 release for prompt-only or moderated workflows.

Note that capability details here draw on Alibaba's API documentation; the "Uncensored" framing and exact release packaging are specific to this catalog listing rather than Alibaba's primary materials.