Creative Studio
Generate, edit, arrange, and reuse images and videos on Runner's project canvas.
Creative Studio gives you a visual canvas for creating campaign and product media from a prompt, selected images, or both.
Entry path
Open a store project → Grow → Creative Studio. The destination opens the Creative Studio canvas.
When the feature is enabled, a generated-image detail view in Runner chat can also offer Open generated image in Creative Studio, which opens the canvas with that output selected.
The image-to-image model picker shows model requirements and funding before Run. Opening the picker does not start generation.
English callouts highlight the controls. Select the image to enlarge it.
An existing example image is selected on the canvas, with prompt, model, and Run controls. Review these inputs before starting generation.
Before you start
- Decide whether you need a new image, an image edit, a new video, or animation from an image.
- Check the selected model’s input, output, quality, limits, and funding label. The picker can show Runner credits or your own Fal key as the funding source; choose the source you intend to use.
- For an edit, add the source image to the canvas and select it first.
- Keep the original asset available until you approve the result.
Generation controls and click results
| Control | What happens after you select it |
|---|---|
| Prompt | Describes the image, edit, motion, or video you want. |
| Model button | Opens Choose a model, with four modality tabs: text-to-image, image-to-image, text-to-video, and image-to-video. |
| Model search and filters | Narrows models by name, generation type, or use case. |
| Star on a model | Adds or removes that model from Favorites. |
| Upload images | Opens the file picker and adds supported images for canvas use. |
| Run | Starts a job with the current prompt, selected model, and selected canvas inputs. |
| Generations | Opens the background queue with running, completed, failed, and interrupted jobs. |
| Clear completed | Removes completed queue entries from the queue view; it does not delete saved media. |
Canvas and asset controls
| Control | What happens after you select it |
|---|---|
| Asset tray | Opens generated assets from the current conversation or all project assets. |
| Asset search | Filters the asset tray. |
| Add to canvas | Places the selected available asset on the canvas. |
| Focus on canvas | Moves the viewport to that asset when it is already present. |
| Media Library | Opens the full saved-media page. |
| Undo / Redo | Reverses or reapplies supported canvas changes. |
| Zoom controls and minimap | Changes the viewport without changing the media itself. |
| Select item(s), then right-click | Opens the canvas context menu. Available actions change with the number and type of selected items. |
| Product tools | For supported image selections, opens Product Photography, Product Scene, Product Placement, Product Holding, Virtual Try-On, Remove or Replace Background, Relight, Material Concept, and Upscale workflows. One-image and two-image tools remain disabled until the required number is selected. |
| Set as product thumbnail | With one image selected, opens the product picker. Choosing Use as thumbnail uploads the image and updates that product’s thumbnail and image list. |
| Extend video / Remove video background | With one video selected, opens the corresponding video workflow when it is available. |
| Export GIF | Converts the selected video to a downloadable GIF; a conversion error leaves the original video unchanged. |
| Duplicate / Crop / Combine images / Layer order / Download | Applies the selected canvas operation when its image count and media type requirements are met. |
| Delete | Removes only the selected item from the canvas. It does not delete the saved asset from Media Library. |
Create media
- Open Creative Studio.
- Choose a model and read its input requirements.
- For text-to-image or text-to-video, leave the canvas selection empty and write a specific prompt.
- For image editing, use Add to canvas on the source image, select it, then choose an Image-to-Image model. For example, Nano Banana 2 Edit appears when available; inspect its requirements rather than assuming every model takes the same inputs.
- For image-to-video, select the single image the model will animate.
- Review the model funding label and current credit notice.
- Select Run once.
- Open Generations to follow progress while continuing other canvas work.
- Inspect the completed output on the canvas. Generated media saves to the media library after successful generation.
Troubleshooting
| State | Meaning and next step |
|---|---|
| Nothing queued | No current background generation is listed. |
| Pending or running | The job is still processing. Keep working and wait for completion. |
| Succeeded | The generated output is available on the canvas and in saved media. Review it before reuse. |
| Failed | The job did not produce the expected output. Read the message, correct inputs, and retry deliberately. |
| Interrupted | Processing ended before success. Confirm funding and provider access before retrying. |
| Run unavailable | Check prompt, selected images, exact input count, model access, and credits. |
| Asset unavailable | Retry loading the tray or choose another saved asset; do not rely on a broken preview. |
Review the setup without starting a job
- Open Choose a model and compare its supported input type with your task.
- Close the picker after selecting an appropriate model.
- Check the prompt and selected images on the canvas.
- Read the funding notice before Run.
Checkpoint: The model, prompt, and source selection describe the intended job. Generation begins only after Run; the queue then provides the status, and a completed output still needs your visual review.
Watch: review an image-edit model
The recording inspects image-to-image requirements and returns to an existing example image on the canvas. It does not start a generation.