Wan 2.7 is an AI video generator powered by the Wan 2.7 engine that transforms text prompts, images, and reference files into cinematic 1080p videos up to 15 seconds long with synchronized audio and consistent characters across multiple shots.
What is Wan 2.7?
Wan 2.7 is the latest version of the AI video generation model from Alibaba Cloud, integrated into the WanOmni all-in-one creation platform. It accepts text prompts, single or multiple reference images (up to 5), video clips (up to 3), and audio files as input, and outputs MP4 (H.264) video files at 30fps in 720p or 1080p resolution. The platform also supports image generation, instruction-based video editing, video continuation, and style transfer.
Key Features
- Cinematic Multi-Shot Videos — Generate multi-scene storyboards from a single sentence; the engine automatically builds scene logic, shot sequencing, and camera language.
- Seamless Audio-Visual Sync — Supports 40+ facial expressions with dialogue generation that matches emotion, lip movement, and voice tone; extends to voice timbre reference via a dedicated parameter.
- 1080p High-Resolution Output — Produces broadcast-ready videos with realistic lighting, cinematic color grading, and consistent quality up to 1080p native resolution.
- Multi-Subject Reference Control — Upload up to five subject references (images, videos, or audio) to lock each character's appearance and voice across multiple shots.
- Professional Camera Control — Supports dozens of camera moves (push, pull, pan, track, crane) plus advanced techniques like Hitchcock zoom and handheld follow, controllable via natural language prompts.
- Instruction-Based Video Editing — Edit existing clips using plain-language commands: remove a bystander, change costume color, swap background season, adjust focus; edited regions blend naturally with the original.
- Visual Assets Generation — AI image generation with precise color palette control, long-text rendering in 12 languages, and ability to fuse up to 9 reference images into a single consistent output.
- New in 2.7 — First and last frame control, 9-grid image-to-video, video continuation, and style transfer (all not supported in Wan 2.6).
Who is it for?
- Marketing & Social Media Managers — Create ads, promos, and social clips by turning product images into animated demos for TikTok and YouTube.
- Film & Animation Pre-Vis Artists — Brainstorm scenes and storyboards with consistent characters; iterate on multi-scene storyboards as if using a virtual studio.
- E-Commerce Merchants — Animate product photos into full-motion showcase videos and generate 360° product demos that boost engagement.
- Educators & Trainers — Create explainer videos and visual lessons with multi-shot storytelling to demonstrate processes step-by-step.
What can you do with Wan 2.7?
- Marketing & Social Media — Quickly produce attention-grabbing ads and promotional clips from text or image inputs, ready for short-form video platforms.
- Film & Animation Pre-Vis — Prototype cutscenes and story-driven content with synced audio, maintaining character consistency across multiple shots.
- E-Commerce Demos — Generate lifestyle clips and product animations that simulate professional studio production without expensive equipment.
- Educational Content — Build visual lessons and training materials using the engine's multi-shot narrative and audio sync capabilities.
How does Wan 2.7 work?
- Upload Your Assets — Upload up to 12 files total (images, videos, audio) as references. You can combine different modalities to express your vision.
- Describe Your Vision — Use natural language to describe the scene. Reference each asset by file type and number (e.g., Image 1, Video 2) in your prompt.
- Generate & Iterate — Generate a 2–15 second video. Extend, edit, or refine by uploading the result and making targeted adjustments through text commands.
