What is Qianwen Creation
Qianwen Creation (formerly Zaodian AI) is a one-stop AI creation platform under Alibaba, powered at its core by Tongyi WanxiangWan 3.0video large models and the Qianwen image model. The platform supports templated creation such as story short films,musicMVs, animatedcomic dramas, and character design, and offers capabilities such as text-to-video, image-to-video, direct document-to-video, and four-modal reference, along with video editing and workflow orchestration.
Main features of Qianwen Creation
-
Story-based creation: Enter a story description on the homepage to generate with one click. It includes four preset templates—story short film, music MV, animated comic drama, and character design—and supports a "director" that automatically breaks down storyboards.
-
Text-to-video: Based on the Wan 3.0 large model, enter a text prompt to directly generate video, with a native maximum of 30 seconds per single generation.
-
Image-to-video / reference-to-video: Upload images or up to 20 reference assets (four modalities: text/image/audio/video) to precisely replicate characters, props, voices, and styles.
-
Direct document-to-video: Supports direct video generation from more than ten formats including doc, ppt, pdf, and md. A single PPT plus one prompt can produce a brand film.
-
Video editing: Edit video like retouching a photo—add or remove elements, replace objects, and switch styles, with local edits requiring no regeneration.
-
Image generation: Based on the Qianwen image model, it supports precise image control and restores naked-eye-level realism.
-
Workflow orchestration: Freely combine nodes such as image generation and video generation to build repeatable creation workflows.
How to use Zaodian AI
- Visit the platform: Visit the Qianwen Creation official website https://create.qianwen.com/ and register or log in to your account.
- Choose a feature: On the official website page, choose the corresponding feature as needed.
- Enter your idea: Enter your creative idea or prompt in the input box. For videos, you can add descriptions of scene transitions, character actions, and more.
- Adjust parameters: Adjust parameters such as stylization, weirdness, and aspect ratio as needed. For videos, options such as intelligent polishing and camera language are supported.
- Generate content: Click the "Generate Now" button, and the AI will generate images or videos based on the input and settings.
- Preview and download: After generation is complete, preview the result. If satisfied, download it to your local device.
Core advantages of Qianwen Creation
-
Wan 3.0 first release across the web: Alibaba's self-developed video large model makes its first release across the web on this platform, natively generating 30-second long videos in a single pass, with image quality limits guaranteed by its own model.
-
Extreme cost-effectiveness: 720P is as low as 0.26 yuan/second, and the API is as low as 0.3 yuan/second, placing it in the industry's top tier for low prices.
-
Four-modal all-around reference: Text, images, audio, and video are input jointly, with up to 20 reference assets at a time, achieving five-dimensional consistency across character/props/voice/scene/style.
-
Direct document-to-video: Directly generate videos from more than ten formats such as doc, ppt, pdf, md, and more. One PPT plus a single prompt is all it takes to produce a brand film.
-
Full closed loop from creation to editing: From story input, automatic storyboarding by the director, and generation to editing capabilities that let you "retouch videos like retouching photos," the entire workflow stays on one platform.
Comparison of similar competing products to Qwen Create
| Dimension | Qwen Create (Alibaba) | Jimeng AI(ByteDance) |
|---|---|---|
| Underlying model | Exclusive self-developed Wan 3.0 + Qwen image, platform-first exclusive | Its own Seedance 2.0 / 2.0 mini / 2.0 VIP + Seedream image model |
| Generation duration | Native 30-second long video, higher one-pass completion rate | Seedance 2.0 is mainly 5-15 seconds, relying on stitching |
| Reference capability | Four-modality (text/image/audio/video) joint reference, up to 20 assets at a time | Supports multi-image reference, but modalities and quantity are relatively limited |
| Signature capabilities | Direct document-to-video (ppt/pdf directly produce videos), Agent Teams multi-agent | Smart canvas, motion imitation, AI lip sync, digital humans, integrated with the CapCut ecosystem |
| Video editing | Independent video editing module, edit videos like retouching photos | Can be transferred to CapCut/CapCut for post-production, with a more mature ecosystem |
Application scenarios of Qwen Create
-
Artistic creation：Artists and designers use AI image generation to create unique artworks or design elements.
-
Advertising and marketing：Marketers use AI to quickly generate attractive advertising images or videos, improving the creativity and appeal of advertisements.
-
Social media content：Social media managers use AI image and video generation to create engaging posts, increasing fan interaction and participation.
-
Education and training：Educators generate teaching materials such as diagrams and animated videos, making learning content more vivid and interesting.
-
Game development：Game developers use AI to generate in-game characters, scenes, or animations, accelerating the development process.
