Make-A-Scene
Meta's AI research tool that generates images from text prompts and freeform sketches
About Make-A-Scene
Make-A-Scene is a text-to-image AI generation system developed by Meta AI Research that combines text prompts with optional freeform sketches for greater creative control. The model uses a scene-based approach with human priors, producing high-fidelity images up to 2048x2048 pixels while enabling scene editing, text editing with anchor scenes, and story illustration generation.
Pros & Cons
Pros
- Combines text and sketch inputs for 66.3% better alignment with creator intent
- High-resolution output up to 2048x2048 pixels
- Scene-based control gives more creative freedom than text-only generators
- Free and open research project from Meta AI
- Supports scene editing and story illustration workflows
Cons
- Research project with no polished consumer-facing interface
- Requires technical expertise to set up and run locally
- No commercial support or SLA available
- Older model (2022) compared to newer alternatives like DALL-E 3 or Midjourney
- Limited community and ecosystem compared to mainstream image generators
Key Features
Text-to-Image Generation
Generates high-fidelity images from natural language text prompts using a transformer-based architecture
Freeform Sketch Input
Accepts optional hand-drawn sketches alongside text prompts to give creators more control over composition and layout
Scene-Based Control
Uses scene layouts as an implicit control mechanism, allowing users to specify elements, arrangements, depth, and structure
High-Resolution Output
Produces images up to 2048x2048 pixels, a four-fold increase over initial capabilities
Scene Editing
Enables editing specific parts of a generated scene while preserving the rest of the composition
Story Illustration
Generates coherent sequences of images suitable for illustrating stories or narratives
Domain-Specific Tokenization
Employs specialized knowledge over key image regions like faces and salient objects for improved image quality
Text Editing with Anchor Scenes
Allows modifying the text prompt while keeping the scene layout anchored for consistent variations
Pricing
- Open research project
- Text-to-image generation
- Sketch-based input
- High-resolution output up to 2048x2048
- Scene editing capabilities
Best For
Concept Art with Composition Control
Artists can sketch rough layouts and let the AI fill in detailed imagery, maintaining precise control over scene composition
Story and Comic Illustration
Create coherent image sequences for stories or comics by using scene anchoring to maintain visual consistency across frames
AI Art Research and Experimentation
Researchers and AI enthusiasts can explore scene-based generation techniques and build upon Meta's open research
Rapid Visual Prototyping
Designers can quickly generate visual concepts by combining rough sketches with text descriptions for faster ideation
Similar Tools
Leonardo.ai
AI-powered creative platform for images, art, and video
Midjourney
The AI image generator known for stunning artistic quality
CGDream
Free AI image generator to visualize your ideas
RenderNet
AI character and video generation with unmatched consistency
DALL-E 3
OpenAI's AI image generator built into ChatGPT for effortless creation
Adobe Firefly
Commercially safe AI image generation integrated into the Adobe Creative Cloud