Wan 2.7 Uncensored

• 🏢 Alibaba's Wan video family, an uncensored text-to-video variant
• 🆕 Built on the Wan 2.7 generation, released June 2026
• 👁️ Photorealistic generation with strong subject coherence across frames
• 🎯 First-frame control exposed through the Wan API
• 🔧 Text-prompted clips with instruction-based editing in the wider family
• 💬 Generates video directly from natural-language text prompts
• 🌐 Part of Alibaba's broader Wan and Qwen generative lineup

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 a text-to-video model in Alibaba's Wan family, packaged as an uncensored variant of the Wan 2.7 generation. According to this catalog it was released in June 2026. Like the standard release, it targets photorealistic generation with strong subject coherence across frames, and it sits alongside the Wan 2.7 Uncensored image-to-video model in the same uncensored line.

The 2.7 generation is also exposed through Wan 2.7. Alibaba Cloud's Model Studio documents the Wan API for both text-to-video generation and image-to-video with first-frame control, indicating the family supports prompt-driven clips as well as conditioned generation from a starting frame. This uncensored build focuses on the text-to-video pathway.

Within Alibaba's own lineage, it follows the same-family Wan 2.6, released in December 2025, and the earlier Wan 2.5 Preview. The catalog positions these as successive Wan video generations rather than competing lines.

Across the broader catalog, Wan sits among Alibaba's recent generative releases, including the newer HappyHorse 1.1 video line and the Qwen image and language models. Treat third-party figures as indicative rather than official, and verify Alibaba's licensing terms before any commercial deployment.