Generate images
Generate images from a prompt and references, choose a model, and find earlier results.
An Image node makes images from your prompt and its references. You write the prompt, pick a model and click Generate.
Generate an image
Click Generate, or press Ctrl/Cmd + Enter.
While the node works, it shows Generating. If Generate is disabled, point to it to see the reason.
Start from a quick preset
An empty Image node shows presets after Try:, such as Restyle, Animate, Upscale and Remove background. Click one to add a small ready-made setup with a sample image. Replace the sample image with your own, then generate.
Choose a model and settings
- Click the model name in the prompt bar and pick a model. Click More to search all models.
- On some models, click LoRA next to the model to add a LoRA.
- Click the settings button next to the model to change Aspect ratio, Resolution and more.
- Set Number of images next to Generate to make more than one image in one run.
The estimated cost next to Generate changes with your settings. Some models need a membership. See Credits and FAQ.
Use references
An Image node can use other nodes as references, for example a character image. Connect the node, or click the Reference tile in the prompt bar. In the prompt, type @ to say which reference you mean. See Nodes and connections.
Find earlier results
A new run replaces the images in the node. Earlier results stay in Generation history. Select the node and click Generation history in its toolbar, or click Generation history in the dock for the whole canvas.
In the history, you can download a result, or click Add to canvas to put it back as a new node.
To work with several images from one run, see Batch runs and groups. To change an image after you generate it, see Edit images.
Was this page helpful?
🎁 Adopted valid suggestions earn a 5,000 Credit reward!