#Start with one clear image idea
Open Image Studio. In the screenshot below, confirm which character the image is for before you touch prompt categories.
Then build one clear idea: style, subject, setting, outfit, pose, and camera. Start with the detail that matters most — do not rewrite everything at once.
Entry points: Generate Image or chat → Generate Images
The chat route brings the current character context into the studio.Category rail: Default: Style, Outfit, Scene, Pose, Camera
Anime mode uses Outfit, Scene, and Pose. Some options display an Unlock state for non-premium accounts.Build one request, then refine one variable
The editor is intentionally a single prompt. Change the category most responsible for the unwanted result rather than rewriting every selection after each generation.
| Workflow | Required input | What happens next |
|---|---|---|
| Generate an image | Selected character plus a non-empty prompt. | The studio shows processing progress and a queue notice. You can browse while it completes. |
| Use a category option | An active category and option selection. | The selected option changes the combined editable prompt; it does not create a separate request. |
| Reset or Random | No additional input. | Reset clears current prompt choices. Random changes visible options, so inspect the resulting prompt before generating. |
| Make Video | A selected saved image in the gallery. | The image becomes the visual source for the video request and its duration/cost choices. |
#Refine from the result, not from guesswork
Use the category rail in the screenshot below — Style, Outfit, Scene, Pose, Camera — and select the options you need. Then click Generate Image and wait for the result instead of starting duplicates.
When an image is close but not right, change one thing and generate again. Saved results land in the gallery for the next step.
#Review the result and gallery
After generation, the screenshot below shows the result beside your prompt controls and gallery. Select a saved image when you want to revisit it or use it as a video source.
#Create video from a selected image
From a selected gallery image, open AI Video. In the screenshot below, pick Standard or Spicy, set the duration, then click Request Video — the token cost is on the button.
Review the live queue and cost before you confirm.
Image request: Generate Image shows 6 tokens
The live button is disabled until a character and non-empty prompt are available.Video source: Select a saved generated image → Make Video
The video modal offers 5- or 8-second duration choices. Confirm the live cost and queue state before starting.A strong source image and one clear motion idea are a better starting point than a long list of competing visual instructions.
#Use the gallery after you generate
Saved results appear in the gallery in the screenshot below. Click a thumbnail when you want to revisit it or use it as the source for AI Video.
While an image is processing, read the queue state — you can leave and return instead of starting duplicate requests.