Zhishiyun Canvas: Keep Ideas, References and Video Steps Together
When prompts, references, and generated results live on separate pages, it is easy to lose the reason behind each step. The canvas brings them into one visual workspace.

Start with a familiar bottleneck
A short-video idea may start in a note, its reference image may sit in an asset library, and a generated clip may appear somewhere else. When you want to revise one shot, you first have to find its prompt and source again. After a project pauses for a few days, picking up the thread gets harder.
Keep the workflow on one canvas
Zhishiyun's infinite canvas lets you arrange text, image, and video nodes in one project and connect them to show how the work relates. You can trace where an idea began, which reference informed it, and what should happen next, then return to a specific node to make a change.
Turn the idea into something you can revisit
Start with a text node for the shot's purpose, audience, visual focus, or an open question. Break a broad idea into short notes that you can refine one at a time instead of burying every requirement in one long prompt.
You remain in control of the creative decisions. The canvas organizes the work; it does not verify brand claims or automatically turn a sentence into a finished storyboard.
Keep references beside the ideas they support
Add image or video nodes near the relevant notes and connect them to show which reference informs which part of the concept. A teammate—or you, later—can see the rationale without reconstructing it from separate screens.
Prompts, references, and generation nodes appear in the same visual project, making it easier to find the context when you review the work.
Continue from the node that needs work
After checking the inputs and prompt, continue with an image or video node that supports generation. If something needs to change, return to that node. The canvas saves project nodes and connections so you can reopen the workflow later.
Available models, parameters, and generation costs depend on what the workspace currently offers. The canvas helps organize creative work; it does not promise a finished video without review.
A simple way to begin
- Write down the shot's intent
Add a text node with the purpose of the image and details that need to stay accurate.
- Connect reference assets
Add image or video nodes and connect them to show how they inform the idea.
- Continue from the node to revise
Adjust the prompt or parameters on a supported generation node while keeping the surrounding context in the project.