Wingman will allow you to generate images from scratch or off existing images. If your AI provider supports image generation, you will see the canvas edit icon under the chat input.
Once GPT-4o dev tooling is released, this will be supported. For now Google AI Studio with Gemini is currently supported
The canvas has several tools available and should be familiar to most simple paint programs.
Here's a comprehensive explanation of the canvas basics:
Drawing Tools
Selection Tools
Color and Style
Image Management
These tools provide a flexible environment for creative image editing and creation, suitable for various design and visualization needs.
Use the Insert
button, located at the top right to import an existing image onto the canvas.
By default, if you have no objects selected in the canvas, Wingman will use the current viewport
- in other words what is visible on the canvas to use as a basis for the image generation, if there are objects present on the canvas.
You can select objects using the selection tool as well, this will only send those selected objects up for image generation.
When generating images you can provide a set of instructions to guide the AI model to produce the results you want. Be as specific as you want based on how you are using the model.