Wanxiang 2.6 - A new generation of video generation models launched by Alitongyi

What is Wanxiang 2.6?

Wanxiang 2.6 is an advanced video generation model series launched by Alibaba Tongyi, designed for professional film and television production and image creation. The model supports role-playing, allowing users to quickly generate short films with cinematic camerawork and narrative by uploading videos and inputting prompts. Simultaneously, the model boasts powerful features such as audio-visual synchronization, multi-shot generation, and sound-driven capabilities, achieving full sensory consistency between visuals and sound. Wanxiang 2.6 improves image quality, sound effects, and command compliance, with single video lengths up to 15 seconds. It is widely used in advertising design, short drama production, and other fields, providing creators with an efficient and convenient creative experience. The model is now available on Tongyi Wanxiang and Alibaba Cloud Bailian platforms.

Main functions of Wanxiang 2.6

• Role-playing functionAfter users upload videos and enter prompts, the model can quickly generate short films with cinematic camerawork and narrative, allowing users to showcase their acting skills in cinematic visuals.
• Storyboard control function: Convert simple prompts into multi-shot scripts to generate coherent narrative videos, maintaining consistency in key information such as subjects and scenes across multiple shots.
• Audio-visual synchronization and sound-drivenIt supports audio-visual synchronization, and can drive the picture action according to the sound signal to achieve perfect coordination between sound and picture.
• Multi-camera generationThe model can generate videos containing multiple shots, maintaining a high degree of consistency in content, rhythm, and atmosphere during shot transitions.
• Improved picture and sound qualitySingle video clips can be up to 15 seconds long, with significantly improved picture and sound quality, meeting professional film and television requirements.
• Instructions followedTo more accurately understand user commands and generate video content that meets user needs.
• General Video EditingIt supports various visual creation capabilities such as text-to-image generation, image editing, text-to-video generation, image-to-video generation, voice-to-video generation, and motion generation, and is suitable for various scenarios such as advertising design, short drama production, and short video creation.

Technical Principles of Wanxiang 2.6

• Multimodal joint modeling and learningWanxiang 2.6 integrates information from multiple modalities (such as video footage, audio, and text) for joint modeling and learning. The model can simultaneously process temporal information such as subject emotions, postures, multi-angle visual features, and acoustic features such as timbre and speech rate. During the generation phase, these elements serve as reference conditions, controlling the style, emotion, and actions of the generated content to achieve full sensory consistency and transfer between visuals and audio.
• High-level semantic understandingThe model possesses powerful semantic understanding capabilities, enabling it to transform user input (such as text prompts, video clips, etc.) into multi-shot segments with complete storylines and narrative tension. Through high-level semantic understanding, Wanxiang 2.6 can construct video content that is logically and emotionally coherent, maintaining consistency in the core subject, scene layout, and environmental atmosphere during multi-shot transitions.

How to use Universal Phase 2.6

Through all phenomena

• Visit the official websiteVisit the Tongyi Wanxiang official website. Follow the prompts to complete account registration and login.
• Select ModelIn the model selection drop-down box, select General Universal Phase 2.2.
• Select Function ModuleSelect the appropriate function based on your needs.
• Upload materialsIf you wish to use the role-playing feature, please upload a short personal video as basic material. For other features, you will need to upload images, text prompts, or other relevant materials.
• Input prompt wordsEnter relevant prompts in the designated input box. For storyboard control, enter "A love story that begins with the male and female protagonists meeting, has a misunderstanding in the middle, and ends with reconciliation."
• Generate contentClick the "Generate" button, and the model will start generating content based on the uploaded materials and the input prompts.
• View and download resultsOnce generated, view the resulting video or image. If satisfied, download it to your local device for saving or use.

Alibaba Cloud Hundred Refinements

• Access Alibaba Cloud Hundred Refinement PlatformVisit the Alibaba Cloud Bailian official website. Complete account registration and login.
• Found the Wan 2.6 model: Find in the platform's AI modelTongyi Wanxiang 2.6 - Image-based video entry.
• Input creation requirementsOn the model page, enter a text description, upload reference images, audio or video clips, or enter storyboard prompts.
• Set generation parametersChoose the video length (maximum 15 seconds), resolution (720P or 1080P), and other generation options.
• Start generating videoClick the "Generate" button, and the model will generate a video based on the input.
• View and download resultsAfter generation, view the generated video effect and download it to your local computer.

Application scenarios of Wanxiang 2.6

• Short video creationThe model can quickly generate creative short video content to meet the needs of social media publishing and improve creation efficiency.
• Advertising productionUse models to generate personalized advertising videos that accurately convey product information and attract user attention.
• Animation ProductionIt efficiently generates animated videos, reduces production costs, and increases the speed of animation creation.
• Educational VideosTo create engaging instructional videos to support teaching and enhance learning interest and effectiveness.
• Film and television special effectsIt assists in film and television production, generating special effects shots or preview videos, and optimizing the creative process.