Duiyou AI video
Duiyou AI Video is an AI video generation capability launched by Alibaba’s design platform
DuiyouAI video
Core parameters and statistics of Duiyou AI Video
Duiyou AI Video is the video generation capability of Alibaba’s design platform 
Dui-you-ai. The underlying layer is based on Alibaba’s open source Wanxiang 2.2 video generation model. It does not exist as an independent product, but is a component of the AI creation ecology of the Duyou platform - users can complete the full-link creation from "photo generation → video generation → brand design → e-commerce marketing → PPT production" within this platform without switching tools.
| Project | Specifications |
|---|---|
| Product form | Duiyou platform’s built-in AI video generation capability |
| Producer | Alibaba (Duiyou d.design) |
| Underlying model | Alibaba Wanxiang 2.2 (open source video generation model) |
| Generation method | Tusheng video, Vincent video |
| Computing power options | Economic mode (simple tasks), flagship mode (high quality) |
| Platform collaboration | Agent, brand design, e-commerce marketing, poster design, PPT production |
| Benefits for newcomers | Register to receive a gift package for newcomers to create for free |
| Usage form | Web online (requires login to Alibaba/Alipay/WeChat) |
| Target Market | Chinese Market |
Positioning Interpretation: Duiyou AI Video is not a video generation SaaS sold separately, but a video module in the "Alibaba Design Family Bucket". Its core value is that "you don't need to export images → go to other tools to generate videos → then import them back" - after the brand design is finalized, the static design can be converted into dynamic videos within the same platform. New users will receive free credits upon registration, lowering the trial threshold.
Boundary Description: Due to the limitations of Wanxiang 2.2 model capabilities, the generated video duration is short (currently about 5-25 seconds). Complex storyboards, long narratives, and precise camera movement control require professional video production tools.
User and market recognition of Duiyou AI Video
The recognition of Duiyou AI Video mainly comes from the technical endorsement of Alibaba Ecosystem and Wanxiang Model.
Model Endorsement: Based on Alibaba’s open source Wanxiang 2.2 video model, the technology source is clear and more credible than gadgets from unknown sources.
Ecological fit: Relying on the Duyou design platform, it is naturally close to the needs of high-frequency short video materials such as e-commerce and marketing.
Items to be verified: The overall user scale of the Duiyou platform and independent usage data of AI video functions have not been publicly disclosed. The commercial licensing terms of the generated video need to be confirmed after logging in.
The cost advantage of Duiyou AI Video
Duiyou AI Video is not an independent paid product, but part of the points/membership system of Duiyou platform.
| Cost Dimension | Duiyou AI Video | Independent Video Generation Tool | Outsourced Shooting |
|---|---|---|---|
| Starting cost | Free gift pack for newcomers | $0-15/month free quota | $500-5000/item |
| Single marginal cost | Consumption based on platform points | $0.01-0.10/time | Fixed quotation |
| Operation complexity | Very low (one click within the platform) | Medium (requires learning) | High (communication + shooting + post-production) |
| Iteration speed | Second level | Minute level | Day level |
| Ecological bonus | Design full-link cooperation with Duiyou | Independent tool | None |
Three-tier cost analysis:
- C client/individual: The free quota of the gift package for newcomers is enough for experience and light use.
- E-commerce seller/small team: Used in conjunction with Duiyou’s brand design and e-commerce marketing capabilities, the overall cost is lower than purchasing multiple tools separately.
- Enterprise: Enterprises within the Alibaba Cloud ecosystem can obtain batch quotas through Alibaba Cloud resource deductions or business plans.
Main functions of Duiyou AI Video
The capabilities of Duiyou AI Video are organized around "rapidly generating video materials within the Duiyou platform":
- Tusheng Video: Upload static pictures (product pictures, design drafts) to generate dynamic videos. Supports two computing power options: "Economic Mode" (simple tasks, low consumption) and "Flagship Mode" (high quality, high consumption).
- Wensheng Video: Enter text description to directly generate video content. Suitable for scenarios that start with creativity and have no ready-made materials.
- Agent Collaboration: The AI Agent capability of the Duiyou platform can assist in generating and optimizing videos, and supports the completion of video creation tasks through natural language interaction.
- Wanxiang Model Driver: Based on Alibaba’s open source Wanxiang 2.2 model, it is suitable for Chinese scenes and e-commerce material styles.
- Full-link collaboration on the platform: After the brand design is finalized, it can be directly converted into a dynamic video within the same platform without exporting/importing. Ensure visual consistency.
Functional synergy value: From design finalization to video generation, the process is completed in a closed loop within Duiyou, ensuring brand visual consistency and eliminating the time loss of switching between multiple tools.
Duiyou AI Video model and version evolution
| Stage | Approximate time | Key changes |
|---|---|---|
| Launch | ~2025-Q3 | Duiyou AI Video is launched as the video generation capability of the platform, providing basic Tusheng/Wensheng videos |
| Wan2.0 | ~2025-Q4 | Access Wanxiang 2.0 model to improve video generation quality and coherence |
| Wan2.2 | ~2026-Q2 | Upgraded to Wanxiang 2.2, added economic/flagship dual mode, supports Agent collaboration |
| Current | 2026 | Continuous iteration of model capabilities, deep integration with Duyou platform Agent/brand/e-commerce capabilities |
The official unified product version number has not been disclosed, and the version context uses the iterative nodes of the Wanxiang model as the main expression dimension.
The technical advantages of Duiyou AI Video
The technical route of Duiyou AI Video is "open source base model + platform engineering packaging":
- Wanxiang 2.2 Base Model: Alibaba’s latest open source video generation model, which has been specifically optimized for Chinese scene understanding and e-commerce visual style consistency. Open source means that the technology is auditable and self-buildable.
- Dual computing power mode: Economic mode (simple tasks, low consumption) and flagship mode (high quality, high consumption) are available.
- Platform-level Agent Orchestration: In conjunction with Duiyou's AI Agent capabilities, it supports multi-step video creation through natural language.
- Full-link closed-loop: Brand design, e-commerce posters, PPT production, and video generation are completed within the same platform.
Product pricing of Duiyou AI Video
The pricing of Duiyou AI videos is nested in the points/membership system of the Duiyou platform, and there is no independent pricing plan.
| Usage | Suitable objects | Features |
|---|---|---|
| Tusheng Video | Those who have static materials | Upload pictures to generate dynamic videos |
| Vincent Video | Creative starter | Enter description to generate directly |
| Platform collaboration | Duyou ecological users | Cooperate with design capabilities to produce materials |
The typical process is "prepare pictures or descriptions \u2192 generate videos \u2192 preview and iterate \u2192 export for use". For first-time users, it is recommended to use simple materials to verify the generation quality and time limit.
Product pricing of Duiyou AI Video (continued)
The pricing model is subject to the official real-time page. Usually a freemium or subscription system is adopted, basic functions can be used for free, and advanced functions or high-frequency use require payment.
Application scenarios of Duiyou AI video
- E-commerce marketing materials: Convert product images into dynamic displays, and the focus of verification is to generate availability rates.
- Short Video Content: Quickly generate creative clips, focusing on coherence.
- Design Proposal: Use generated video to supplement the static plan, and the focus of verification is consistency with the brand style.
Applicable groups of Duiyou AI Video
- E-commerce and Marketing Team: High-frequency, low-cost dynamic materials are needed.
- Designer: I hope to quickly transform static works into dynamic displays.
- Content Creator: Lack of shooting conditions but need video material.
Unsuitable situations are: professional film and television productions that require long videos, precise storyboard control, or extremely high picture continuity requirements.
Summary and outlook of Duiyou AI Video
The core value of Duiyou AI Video is to connect Tusheng/Wensheng video capabilities to the Alibaba Wanxiang model and integrate it into the design platform to provide rapid dynamic material output for e-commerce and marketing. The clear source of the model and its adaptability to Chinese scenes are its practical advantages.
The current limitation is that AI video generation still has common shortcomings in the industry in terms of duration, consistency and controllability, and public pricing is not transparent. For teams, it is recommended to use the free quota to verify the availability of common materials before deciding whether to pay for expansion; the authorization scope and duration limit of the generated content should be confirmed before commercial use.
Related tools: runway, pika
How to use Duiyou AI Video
- Web client: You can use it by visiting the official website and registering an account. Most functions do not require installation.
- API access: Provides RESTful API, developers can obtain the API Key and integrate it into their own applications.
Version Info
- Duiyou AI video (Wanxiang 2.2 model) ：Currently based on Alibaba Wanxiang 2.2 video generation model, it supports Tusheng Video and Wensheng Video, provides dual computing power options of economic mode and flagship mode, and collaborates with Duyou platform Agent/brand design/e-commerce marketing capabilities.
- AI video capabilities are online ：Duiyou AI Video is launched as the video generation capability of Duiyou AI, providing basic Tuxing videos and Wensheng videos. There is no official precise date yet, it is recorded according to the public collection time.
User Reviews