Wan2.7-Video - Alibaba Tongyi's AI Video Creation Model

What is Wan2.7-Video

Wan2.7-Video is launched by Alibaba Tongyi LabAI videocreation large model, supporting full-modal input of text, images, video, and audio. The model breaks through traditional generation limitations, enabling full-process creation such as local editing "like retouching a photo but for video", dialogue and action adjustment, camera movement replication, and plot continuation. Wan2.7-Video supports control of 5 main characters and multi-grid storyboards, driven by the "drama core" for professional storyboarding, 40+ types of expressions, and cinematic camera movements, greatly lowering the threshold for professional video creation.

Main Features of Wan2.7-Video

• Precise local editing: Users can make local adjustments to the video image through instructions, such as adding or removing elements, replacing objects, or modifying attributes, and the lighting and materials after editing can naturally blend with the original video.
• Environment style transformation: While keeping the character's actions unchanged, the system supports one-click switching of background seasons or converting the image into diverse artistic styles such as wool felt and cyberpunk.
• Plot shooting modification: Without starting over, users can adjust character dialogue and match lip sync and emotions, modify behavioral actions, or change camera positions and shot types through instructions.
• Plot continuation control: Through the continuation function combining the first and last frames, the model precisely controls the image structure while preserving the natural continuation of video dynamics, achieving seamless connection and extension of the plot.
• Rapid creative replication: The system supports retaining the original video's action sequence, camera movement, or style effects and applying them to new scenes, enabling one-click rapid reuse of dynamic creativity.
• Multi-subject character control: Supports uploading image, video, and audio references for up to 5 subjects, precisely locking character appearance and exclusive voice, ensuring high consistency of features across multiple shots.
• Storyboard shot control: Through multi-grid reference images, users can precisely control the story direction, camera composition, and character settings, achieving storyboard-level precise execution of shots.
• Intelligent script generation: Based on deep learning of professional scripts, the model can automatically generate dramatic logic-compliant narrative structure and professional storyboard scripts from a one-sentence idea from the user.
• Cinematic style control: Driven by the drama core, it directly generates corresponding lighting, cinematography, and color parameters, supporting free combination of thousands of cinematic styles and maintaining consistency across multiple shots.
• Professional camera movement execution: The system can precisely execute complex composite camera movement techniques such as Hitchcock zoom and rise reveal.
• Delicate expressions and voice: Supports performing over 40 subdivided expressions, with generated dialogue accurate and voice vivid and natural, achieving textured professional performances through audio-visual synchronization technology.

How to use Wan2.7-Video

• Alibaba Cloud Bailian: VisitAlibaba Cloud Bailianto enter the model plaza, select the Wan series models to call the API or experience on the web.
• Wanxiang Official Website: VisitTongyi WanxiangThe official website provides a visual operation interface, supporting direct upload of materials for creation.
• Usage: Supports full-modal input of text, images, video, and audio, controlling画面 structure, plot direction, local details, and temporal changes through natural language instructions, enabling full-process creation such as generation, editing, replication, and continuation.

Key Information and Usage Requirements of Wan2.7-Video

• Development Team: Alibaba Tongyi Lab
• Product Positioning: An AI video creation suite covering the full process of generation, editing, replication, continuation, and reshaping
• Input Modalities: Supports full-modal input of text, images, video, and audio
• Subject Control: Supports up to 5 subjects, can lock appearance features and专属 timbre, maintaining multi-shot consistency
• Core Capabilities: Precise local editing, plot/dialogue/camera position modification, action and camera movement replication, plot continuation, storyboard shot control
• Performance Capabilities: Supports 40+ subdivided expressions, accurate dialogue generation, natural voice with audio-visual synchronization
• Camera Movement Support: Dozens of basic camera movements (push, pull, pan, track, etc.) and composite camera movements (Hitchcock zoom, rising reveal, and other cinematic techniques)
• Access Channels: Alibaba Cloud Bailian or Wanxiang official website
• Operation Method: Natural language instruction control, no programming background required

Core Advantages of Wan2.7-Video

• Full-Modal Input Fusion: Supports any combination of text, images, video, and audio input, enabling comprehensive control over image structure, plot direction, local details, and temporal changes.
• Full-Process Creation Coverage: From video generation to local editing, creative replication, plot continuation, and character reshaping, it provides a complete toolset spanning the entire creation chain, with no need to switch between multiple platforms.
• Precise Local Editing: Breaks through the traditional regeneration mode, supports instruction-level addition and removal of elements, object replacement, and attribute modification; the lighting and materials of the edited area blend naturally with the original video, enabling video editing like photo retouching.
• Controllable Filming and Plot: No need to start over; you can adjust character dialogue (automatically matching lip sync and timbre), modify actions and behaviors, and change camera positions and shot types through instructions, enabling flexible secondary creation.
• Multi-Subject Consistency: Supports locking the appearance and timbre of up to 5 subjects, ensuring that the same character maintains highly consistent features across multiple shots, with each character having a专属 voice performance.

Comparison of Wan2.7-Video with Similar Competitors

Application Scenarios of Wan2.7-Video

• Film and Television Content Creation: Low-cost production of independent films, short films, and animations, quickly visualizing scripts through storyboard storyboarding, or for dynamic previsualization and shot testing before formal shooting.
• Short Videos and Social Media: Creators quickly generate plot-based, costume-change, and special-effects short videos, supporting replication of popular camera movements and multi-character plot performances, adapting to the content needs of platforms such as Douyin, Kuaishou, and Instagram.
• Advertising and E-commerce Marketing: Rapid generation and iteration of product display videos, supporting local editing to replace product elements, adjust shooting camera positions, as well as multi-angle display and voice-over dubbing for virtual models.
• Education and Training: Produce teaching demonstration videos, historical scene restoration, and visualization of experimental processes, building coherent knowledge explanation sequences through the plot continuation feature.
• Music and Entertainment: Achieve specific stylized visuals in MV production (such as wool felt, cyberpunk), dance motion replication and camera movement references, as well as multi-shot consistent performances by virtual singers.