Figma Make brings AI-powered design generation directly into Figma. Describe what you want in natural language, and Make generates designs on your canvas. It's not a separate tool - it works within your existing Figma files, using your components and styles. This means AI-generated designs are immediately part of your workflow, not exports that need importing.
1. Accessing and Activating Figma Make Features
Getting started with Make in your Figma workflow:
Enabling Make
Figma Make is available in Figma's AI features. Access it from the toolbar or use the keyboard shortcut. It opens as a panel where you can type prompts and see results on your canvas.
Understanding the Interface
The Make panel shows your prompt input, generation options, and history. Generated designs appear directly on your canvas as editable Figma elements - not flattened images.
Project Context
Make understands your current file. It sees your color styles, components, and design patterns. When you ask for a button, it can use your existing button component if available.
Generation Settings
Control generation parameters: how closely to match your existing styles, level of creativity, and whether to use existing components. Adjust based on whether you want consistency or exploration.
2. Generating UI Designs from Text Prompts
Write prompts that generate exactly what you need:
Describing Components
Be specific about what you want: 'A pricing card with three tiers, the middle one highlighted, each showing price, features list, and a CTA button. Use our brand colors.' Clear descriptions get accurate results.
Layout Specifications
Include layout details: 'A dashboard header with logo on the left, navigation links in the center, and user avatar with dropdown on the right. Fixed at the top with a subtle shadow.'
Style References
Reference styles: 'Modern and minimal like Linear', 'Playful with rounded corners like Notion', 'Clean enterprise style with plenty of whitespace'. Make understands these references.
Context and Purpose
Explain the context: 'An onboarding screen for a fitness app where users select their goals. Friendly and motivating tone. Users can select multiple options.' Purpose guides design decisions.
3. Iterating and Refining AI-Generated Designs
First generation is your starting point, not your destination:
Follow-Up Prompts
Refine with additional prompts: 'Make the buttons larger', 'Add more spacing between sections', 'Use a gradient background instead'. Each prompt builds on the previous result.
Selective Regeneration
Select specific parts and regenerate just those. Keep the header you like, regenerate the content area. Mix and match generations for the best result.
Manual Refinement
Generated designs are fully editable Figma elements. Adjust spacing, change colors, modify text - all normal Figma operations work. AI generation plus manual polish is the optimal workflow.
Variation Generation
Ask for variations: 'Show me three different approaches to this layout' or 'Generate a dark mode version'. Compare options before committing to a direction.
4. Combining AI Generation with Manual Design Work
AI and manual design work best together:
Starting with AI
Use Make to generate initial concepts quickly. Explore different directions in minutes instead of hours. Find the right approach faster, then refine manually.
AI for Repetitive Elements
Generate repetitive elements: 'Create 10 different user cards with varied content for a testimonials section'. AI handles the variations, you curate the results.
Manual Design, AI Enhancement
Design the core elements manually, then use Make to extend them. 'Based on this card I designed, create variations for different content types.' AI learns from your work.
Component Integration
Convert AI-generated designs into components. Add them to your design system. Future generations can use these components, building on your curated library.
Key Takeaways
- Figma Make generates editable designs directly on your canvas
- Write detailed prompts with layout, style, and context specifications
- Iterate through follow-up prompts and selective regeneration
- Combine AI generation with manual refinement for best results
- Generated designs can become reusable components in your system