dicebear/dicebear
Dicebear
DiceBear is an SVG avatar generation library and image API that creates deterministic profile pictures from seed strings. It provides a system for generating consistent visual outputs across different environments and programming languages using JSON style definitions.
The project distinguishes itself with a comprehensive design toolkit, including a visual style designer and a Figma integration plugin that converts design frames into JSON schemas. It supports sophisticated visual control through weighted probability distributions for components, contrast-aware color management for accessibility, and a transformation engine for randomized rotation and scaling.
Beyond basic generation, the system covers a broad range of capabilities including multi-format exports to raster images like PNG and WebP, a command line interface for batch asset production, and the ability to host a self-managed API for remote rendering. It also includes security features such as SVG output sanitization to prevent script injection.
Features
- Seeded Generation - Transforms seed strings into repeatable random values to ensure consistent visual outputs across environments.
- Visual Determinism - Creates consistent profile pictures from seed strings to ensure the same input always produces the same visual result.
- Deterministic Image APIs - Serves seed-based avatar images and style metadata via an HTTP service without requiring local installations.
- Local Vector Generation - Enables the generation of consistent SVG avatars directly within local application code.
- Remote Image Generation - Enables the creation of consistent avatar images by passing seed strings and parameters through HTTP request URLs.
- SVG Element Generators - Provides a library that generates consistent, deterministic profile pictures from seed strings using SVG and JSON.
- Deterministic - Generates independent, deterministic attribute values using keys to ensure consistency regardless of call order.
- Cross-Language Visual Consistency - Ensures that the same seed produces byte-identical avatar images regardless of the programming language used.
- Avatar Style Schemas - Provides a declarative JSON schema for defining canvas size, elements, and color palettes for avatar generation.
- Visual Style Schemas - Uses a structured JSON format to define canvas elements, visual components, and probability weights.
- Visual Variant Weighting - Allows adjusting avatar appearance through style-specific options and weight maps to prefer or exclude variants.
- Deterministic Attribute Derivation - Generates independent values for visual traits using keys to maintain consistency regardless of call order.
- Avatar Appearance Configurators - Enables the definition of visual styles and component probabilities to create unique sets of generated characters.
- Probabilistic Component Sets - Uses a system of visual variants and probability ranges to determine avatar looks based on a seed.
- Declarative Style Objects - Uses JSON definitions and map-based descriptors to declaratively configure avatar styles and options.
- SVG Export Pipelines - Produces and exports scalable vector graphics in various formats for web and mobile user interfaces.
- User Avatars - Generates unique placeholder images and avatars for users via a local library or HTTP API.
- Raster Image Exports - Transforms SVG images into raster formats such as PNG, JPEG, WebP, and AVIF.
- Self-Hosted API Services - Allows the deployment of private instances of the avatar generation service to control privacy and rate limits.
- Vector Rasterizers - Transforms vector-based XML output into multiple raster formats like PNG and WebP.
- Remote Rendering Servers - Produces images on a remote server via query parameters to provide assets without local libraries.
- SVG Image Exports - Provides the ability to export generated avatars as XML-formatted SVG strings or data URIs.
- Multi-Format Asset Exports - Outputs generated images as XML strings, JSON objects, or data URIs for different application contexts.
- Rendering Randomizers - Applies randomized rotation, translation, and scaling to avatar components within defined ranges.
- Visual Variant Probabilities - Determines the frequency of specific visual components in outputs using weights and probability percentages.
- Weighted Component Distribution - Assigns numerical weights to visual variants to control how frequently specific components appear.
- Automatic Contrast Adjusters - Dynamically computes and applies colors using contrast sorting to ensure visual accessibility.
- Dynamic Text Interpolation - Allows the insertion of seed-derived initials or user-defined font settings into text elements.
- Figma to JSON Style Conversion - Includes a Figma plugin that converts design frames into JSON style definitions, including probabilities.
- Color Palette Management - Implements a system for resolving colors and managing palettes using contrast-based sorting for accessibility.
- Visual Element Transforms - Provides a system to adjust the position, scale, rotation, and flip of avatar components.
- Visual Style Customization - Provides a graphical interface for browsing styles and modifying appearance options without writing code.
- Style Editors - Ships a visual editor for creating custom style definitions and exporting them as JSON files.
- Remote Avatar Embedding - Allows fetching and embedding avatar images via HTTP URLs in any environment without requiring local installations.
- Server-Side Rendering - Supports generating avatars on the server to eliminate client-side JavaScript and hydration issues.
- Image Generation API Hosting - Provides a standalone server for exposing the avatar generation pipeline as a remote API.
Star history
How this analysis was created: This summary and feature list were written by an AI model that read the project's README and public documentation pages. Each feature links to the documentation it came from; stars, license and language come straight from the GitHub API. The model does not read the source code, and the analysis is refreshed when the project is re-analysed. Learn more on our About page.
