Your First Result

You will build one text prompt that becomes one generated image on the canvas.

Create the prompt

1. Open any .canvas file in Obsidian, or create a new one.
2. Double-click empty space to create a text card.
3. Type a short prompt. For example:

A small wooden cabin beside a quiet lake at sunrise, soft light, realistic photo.

Run the model

1. Select the text card. The floating canvas toolbar shows Bragi buttons.
2. Click Generate image.
3. A generate bar appears below the selection. Confirm:
   • An image model is selected. Only models from configured providers appear.
   • Parameters such as Aspect Ratio if shown.
   • Count is set to x1.
4. Click Run.

What you should see

1. A placeholder card appears beside your prompt with a loading animation.
2. When the request finishes, the placeholder becomes an image file card.
3. An edge connects the prompt to the result.
4. The image file is saved in _bragi/assets/ at the root of your vault. See Where Files Go.

If no model appears

You need a provider key and at least one enabled image model. See No Models Listed.

Try text next (optional)

Select a text node → Generate text → Run. Ask the model to separate five short prompts with a line containing only ---SPLIT--- on its own line. Bragi can split the answer into multiple text cards. This workflow is covered in Split Text into Batch Images.

Next steps

• Make Five Storyboard Frames
• Canvas Basics