Wan2.7-Video - An AI video creation model launched by Alitongyi

What is Wan2.7-Video?

Wan2.7-Video is an AI video creation model launched by Alibaba's Tongyi Lab, supporting full-modal input including text, images, video, and audio. Breaking through traditional generation limitations, the model enables a complete creation process, allowing for "video editing like photo editing," including partial editing, dialogue and action adjustments, camera movement replication, and story continuation. Wan2.7-Video supports control of five main characters and multi-grid storyboards, using a "drama core" to drive professional storyboarding, 40+ facial expressions, and cinematic camera movements, significantly lowering the barrier to entry for professional video creation.

Main functions of Wan2.7-Video

• Precise local editingUsers can make local adjustments to the video frame by adding or deleting elements, replacing objects, or modifying attributes through commands, and the edited lighting and materials can be naturally integrated with the original video.
• Environmental style changeWhile keeping the character's movements unchanged, the system supports one-click switching of the background season or conversion of the image to diverse art styles such as felted wool or cyberpunk.
• Filming modificationsNo need to start over; users can adjust character dialogue and match lip movements and emotions, modify actions, or change camera angles and shot types simply by giving commands.
• Plot continuation controlBy combining the first and last frames with the continuation function, the model can accurately control the screen structure while preserving the natural continuity of the video's dynamics, thus achieving seamless connection and extension of the plot.
• Creative quick replicationThe system supports preserving the original video's motion sequences, camera movements, or style effects, and applying them to new scenes, enabling one-click quick reuse of dynamic creative ideas.
• Multi-agent role controlIt supports uploading images, videos, and audio references for up to 5 subjects, accurately locking in the character's appearance and unique voice, and ensuring a high degree of consistency in features across multiple shots.
• Storyboard controlWith the help of multi-grid reference images, users can accurately control the story direction, camera angles, and character settings, and achieve precise storyboard-level execution of the storyboard.
• Intelligent script generationBased on deep learning of professional scripts, the model can automatically generate a dramatic and logically coherent storyboard based on a user's single creative idea.
• Cinematic Style ControlDriven by the core of the scene, it directly generates corresponding lighting, photography, and color parameters, supporting the free combination of thousands of cinematic styles and maintaining consistency across multiple lenses.
• Professional camera operationThe system can accurately execute complex composite camera techniques such as Hitchcock zoom and rising revelation.
• Delicate facial expressions and voiceIt supports the expression of more than 40 subdivided expressions, generates accurate lines and vivid and natural sounds, and achieves a high-quality professional performance through audio-visual synchronization technology.

How to use Wan2.7-Video

• Alibaba Cloud Hundred RefinementsVisit Alibaba Cloud Bailian to enter the Model Plaza, select the Wan series model to call the API or experience it on the web.
• Wanxiang Official WebsiteVisiting the Tongyi Wanxiang official website provides a visual interface that supports direct uploading of materials for creation.
• How to useIt supports full-modal input of text, images, video, and audio, and controls the screen structure, plot direction, local details and temporal changes through natural language commands, realizing the entire creation process such as generation, editing, replication and continuation.

Key information and usage requirements for Wan2.7-Video

• Development TeamAlibaba Tongyi Lab
• Product PositioningAn AI video creation suite covering the entire process of generation, editing, replication, continuation, and reshaping.
• Input modeSupports full-modal input including text, images, video, and audio.
• Main controlSupports up to 5 subjects, with the ability to lock facial features and unique voice to maintain consistency across multiple cameras.
• Core CompetenciesPrecise local editing, plot/dialogue/camera position modification, motion camera movement replication, plot continuation, storyboard panel control.
• Performance abilitySupports 40+ subdivided facial expressions, accurate dialogue generation, natural sound, and synchronized audio and video.
• Camera movement supportDozens of basic camera movements (push, pull, pan, tilt, etc.) and compound camera movements (Hitchcock zoom, rising reveal, and other cinematic techniques).
• Access ChannelsAlibaba Cloud Bailian or Wanxiang official website
• Operation methodNatural language command control, no programming knowledge required.

Wan2.7-Video's core advantages

• Multimodal input fusionIt supports any combination of text, images, video, and audio input, enabling comprehensive control over screen structure, plot development, local details, and temporal changes.
• Full-process creation coverageFrom video generation to partial editing, creative replication, story continuation, and character reshaping, it provides a complete toolset that spans the entire creative process, eliminating the need to switch between multiple platforms.
• Precise local editingBreaking through the traditional regeneration mode, it supports adding and deleting elements, replacing objects, and modifying attributes at the command level. The lighting and materials of the editing area are naturally integrated with the original video, making video editing as easy as photo editing.
• Filming plot is controllableWithout having to start over, you can adjust character dialogue (automatically matching lip movements and timbre), modify actions, and change camera angles and shot types via commands, enabling flexible secondary creation.
• Multi-subject consistencyIt supports locking the appearance and voice of up to 5 subjects, ensuring that the same character has highly consistent characteristics across multiple shots, and each character has its own unique voice performance.

Comparison of Wan2.7-Video with similar competing products

Application scenarios of Wan2.7-Video

• Film and television content creationLow-budget production of independent films, short films, and animations can quickly visualize the script through storyboards or be used for dynamic pre-production and shot testing before formal shooting.
• Short videos and social mediaCreators can quickly generate short videos in the form of storylines, costumes, and special effects, supporting the replication of popular camera movements and multi-character storytelling, and adapting to the content needs of platforms such as Douyin, Kuaishou, and Instagram.
• Advertising and e-commerce marketingThe ability to quickly generate and iterate product display videos, supporting partial editing to replace product elements, adjust camera positions, and provide multi-angle virtual model displays and voice-over narration.
• Education and training sectorIt can create teaching demonstration videos, restore historical scenes, and visualize experimental processes, and build a coherent sequence of knowledge explanations through the plot continuation function.
• Music and EntertainmentThe music video production process achieves specific stylized visuals (such as felted wool and cyberpunk), replicates dance moves and provides camera movement references, and ensures consistent performance of virtual singers across multiple camera angles.