Core Summary
On August 6th, Alibaba's video generation model Wan3.0 launched a public beta test. This upgrade has transformed it from a tool that can simply generate videos into one that can actually assist users in their tasks: it can now create complete 30-second videos, convert documents such as PPTs and Word files directly into videos, significantly improve the realism of the generated images (each character appears unique with distinct features), and also allows for video content editing. The public beta channels and API prices have been made available, enabling both individuals and businesses to utilize this technology quickly.
Detailed Explanation
1. **30-second Videos: From “Taking a Segment” to “Telling a Complete Story”**
Previously, AI-generated videos were only a few seconds long and could only display a single action (for example, someone waving their hand). Now, Wan3.0 can produce 30-second videos that are long enough to tell a brief story. For instance, if you want to create a video on “how to measure heart rate using a smartwatch,” the 30 seconds could include the entire process from picking up the watch to performing the measurement and displaying the heart rate results. The video can also feature smooth transitions between shots (from the watch to the wrist to the screen) or use a single continuous shot, making it more natural and engaging. Additionally, the model includes an “intelligent duration” function that automatically recommends the appropriate length based on your input, eliminating the need to worry about whether to make the video 15 seconds or 20 seconds.
2. **Converting Documents into Videos: One Click to Create Presentations or Product Promos**
This is one of the most practical improvements. Wan3.0 can directly read common document formats such as DOC, XLS, PPT, and PDF files. For example, if you are a salesperson with a PPT about smart glasses, you can upload it and specify a prompt like “Generate a vivid product promotional video that highlights the lightweight design and AR features.” The model will then convert the images and information from the document into a video that includes visual demonstrations of the glasses, explanations of their functions, and even dynamic effects. This process used to take several hours of editing, but now it can be done in just minutes, saving a lot of time for professionals (each file is limited to 100MB/50 pages, which covers most everyday documents).
3. **Enhanced Realism: Unique Characters with Natural Expressions**
Previously, AI-generated characters often looked similar or had stiff expressions. Wan3.0 has addressed this issue by creating characters with unique features and realistic expressions. Each character has distinct facial features and skin textures (for example, some have square faces, while others have freckles), and their emotions are expressed naturally (frowning when angry, smiling with slight wrinkles at the corners of the mouth). The body movements and expressions are also well-coordinated (for instance, hands may move naturally while speaking). Even the charts are professionally designed, making the videos look as if they were produced by a professional team.
4. **From a “Toy” to a “Tool”: A Productive Assistant for Professionals**
Previously, AI-generated videos might have been used just for entertainment (such as creating funny clips), but now Wan3.0 has become a useful tool. Teachers can use it to create animated lesson plans, marketers to produce promotional videos for products, and business professionals to make PPTs more engaging when presenting. The model also supports editing, allowing you to modify unsatisfactory parts of the video or adjust the storyline and dialogue. However, there are some minor drawbacks: the sound quality is somewhat mechanical, and the text may sometimes appear incorrect, but overall, it is sufficient for everyday use.
5. **How to Use It? How Much Does It Cost?**
If you want to try out Wan3.0, you can access it through platforms such as Alibaba Cloud Bailian, Wanjing Yike, Wanxiang’s official website, Qianwen Creation (PC version), IF STUDIO, and Duoyou. The Qianwen APP is currently available for a limited number of users in a beta phase. If you wish to integrate the model into your own applications, the API prices are as follows: 0.3 yuan per second for 480P resolution, 0.6 yuan per second for 720P resolution, and 1.2 yuan per second for 1080P resolution. For example, generating a 30-second 1080P video would cost 36 yuan, which is not expensive for small teams or individual creators. The API will be fully available in the near future.
This upgrade of Wan3.0 represents a significant shift from using AI for entertainment to using it as a tool that enhances productivity. It enables ordinary users to create professional-quality videos quickly, whether for business presentations, product promotions, or educational content.