Skip to main content
The prompt panel is where image generation and natural language editing begin in Recraft Studio. It provides access to generation modes, models, styles, and settings, and adapts based on whether you are creating new content or editing an existing selection.

​
What the prompt panel controls

The prompt panel handles all prompt-driven actions, including:
  • generating new images, frames, mockups, and image sets
  • editing selected images using natural language
  • selecting models, styles, and generation modes
  • configuring generation parameters such as aspect ratio and number of images
It does not manage layout, selection, or object-specific actions such as alignment, arrangement, or export. Those are handled on the canvas and in the context panel.

​
Prompt input and controls

The prompt panel includes controls for configuring how images are generated or edited:
  • Prompt field: Enter natural language descriptions or instructions
  • Generation mode: Choose how prompts are interpreted (for example, Agentic or Manual)
  • Model selection: Select the model used for generation or editing
  • Style selection: Apply visual styles to generation
  • Aspect ratio: Control image dimensions
  • Images per generation: Generate multiple outputs at once
Some controls are available only in specific modes or when editing an existing selection.

​
Prompt panel states

The prompt panel changes behavior based on selection state. No selection When nothing is selected, the prompt panel is in generation mode. You can:
  • enter a prompt to generate new images
  • choose a model and style
  • set aspect ratio and number of images
  • start new generations
This is the default state for creating new content. Single object selected When a single image or object is selected, the prompt panel switches to editing mode. You can:
  • modify the prompt associated with the selected image
  • apply natural language edits or refinements
  • regenerate variations based on the current selection
This state is used for iterating on existing images rather than creating new ones. Multiple objects selected When multiple images are selected, the prompt panel switches to reference mode. In this state:
  • the selected images are treated as reference images for a new generation
  • any prompt you enter is interpreted in relation to the selected images
  • the next generation creates a new image informed by those references, rather than editing them directly
This state is used to generate new content guided by visual examples on the canvas.

​
Docking and placement

The prompt panel can be toggled between different positions depending on your workflow:
  • docked at the bottom of the canvas
  • docked to the side of the workspace using the prompt panel toggle
Docking affects layout only. All prompt panel functionality remains the same regardless of position.