- Wan3.0 lets users upload presentations, spreadsheets and PDFs and generate videos directly from them
- Alibaba doubles Wan3.0’s maximum video length while targeting greater consistency, realism and production-ready output
Alibaba Group’s new video-generation model can turn a PowerPoint presentation, Excel spreadsheet or PDF into a finished video, pushing AI video tools beyond text-to-video prompts and deeper into everyday business workflows.
Wan3.0, officially released August 24, accepts documents of up to 100 megabytes and 50 pages, allowing users to upload presentation decks, reports or other business materials and have the model interpret their structure and generate video content without writing additional prompts.
Widely seen as a rival to ByteDance’s Seedance and Kuaishou’s Kling, the model also doubles the maximum length of a single generated video to 30 seconds from 15 seconds, enabling longer sequences and more complex camera movements.
Beyond documents, Wan3.0 supports text, images, audio and video as inputs. Alibaba says it has improved consistency across characters, objects, voices, spatial relationships and visual styles.
Maintaining continuity
This helps to address a persistent problem with AI-generated video: maintaining continuity across a sequence rather than producing a collection of visually appealing but disconnected shots.
The company is also targeting more realistic human imagery, with improvements to facial features, skin texture and subtle expressions.
Wan3.0 entered public testing on August 6 and has already been used in commercial workflows including short dramas, advertising and tourism content.
It is now available through Alibaba Cloud’s Bailian platform, the Wanxiang website and the Qwen app.
API pricing starts at 0.3 yuan ($0.0045) per second for 480p video, 0.6 yuan for 720p and 1.2 yuan for 1,080p. Alibaba is offering a 30% promotional discount from August 24 through September 23.
Why it matters globally
The bigger shift may be less about video quality than the starting point.
By allowing users to begin with a presentation or spreadsheet rather than a blank text prompt, Alibaba is positioning generative video as a business productivity tool as much as a creative one.
For international users, Wan3.0 highlights the direction of competition in AI video: as generation quality becomes increasingly commoditized, the battle may increasingly be fought over workflow integration, input formats and the distance between an idea and a finished video.
