Extract prompt lets you generate a descriptive text prompt from any image, whether it was made in Recraft or uploaded from elsewhere. The tool analyzes the visual content and returns a detailed description of what it “sees,” including aspects like subject, composition, technique, and style.This is especially useful when you want to:
Understand how to describe an existing image in prompt form
Recreate or iterate on a similar image
Explore stylistic attributes for reference or learning
Reuse successful results in new contexts
Refine previous generations without starting from scratch
Analyze how a particular outcome or style was achieved
Unlike reusing a generation’s original prompt, Extract prompt works as an AI interpretation. It does not recover the exact text that was originally used to generate the image.
Upload an image from your computer or select one on your canvas.
Click Extract prompt under the prompt field.
Recraft will analyze the image and display a text description summarizing its content and visual style.
You can edit this description or use it directly as a new prompt for generation.
Tip: Extract prompt works best with clear, well-defined visuals (illustrations, product photos, icons, etc.). The richer the image detail, the more specific and useful the extracted prompt will be.
