Map of GitHub
Every dot is a repository. Every cluster is an ecosystem. Navigate 700,000+ open-source projects mapped by co-star similarity — find related tools, discover new libraries, and see the structure of the open-source world at a glance. The same open-source Vue/MapLibre app is built into this site (not embedded in an iframe).
Loading map…
Side panels in the explorer use viewport-fixed positioning inside the upstream UI; site navigation stays above the embed via a higher stacking context. Scroll the page normally — the galaxy surface itself is draggable inside the frame area.
How to Use the Map of GitHub
Basic Navigation
The map opens with a full view of the GitHub galaxy. Scroll to zoom in — repository names become visible as you get closer. Click and drag to pan. On mobile, pinch to zoom and swipe to pan. The map contains hundreds of thousands of points, so rendering may take a moment when first loading.
Searching for a Repository
Use the search bar to find any specific repository by name. Type the repository name (with or without the owner prefix,
e.g. react or facebook/react). The map will zoom to that repository location and highlight it.
Look at nearby repositories — those are the projects most commonly starred by the same developers.
Exploring Clusters
Zoom into any dense region to reveal its structure. Each major language and framework ecosystem forms its own neighborhood. Within those neighborhoods, sub-clusters form around specific use cases — testing tools, UI frameworks, build systems, data processing libraries. The map makes the invisible architecture of the open-source ecosystem visible.
Clicking on Repositories
Click any dot on the map to see the repository name and a link to its GitHub page. From there you can star it, read the README, or clone it. The map acts as a visual discovery layer on top of GitHub full functionality.
How the Map of GitHub Works
The Map of GitHub is built from a massive dataset of GitHub star relationships — records of which developers have starred which repositories. Two repositories that are frequently starred by the same developers are considered similar and placed near each other on the map.
This co-star signal is fundamentally different from tag-based similarity. A repository tags are chosen by its author; co-star similarity is determined by the collective behavior of the developer community. The result is a map where proximity reflects real-world usage patterns — tools that developers actually use together end up as neighbors, regardless of how they are categorized.
The edition of the Map of GitHub here covers hundreds of thousands of repositories that passed quality filters, computed from massive star relationship graphs across millions of GitHub accounts.
Notable Clusters on the Map
JavaScript & Node.js
The largest single region on the map. The JavaScript ecosystem — npm packages, frontend frameworks, build tools, and Node.js server libraries — forms a densely interconnected galaxy of its own. React, Vue, Angular, and their respective ecosystems are clearly separated as sub-clusters within the larger JS region, but all remain close enough to show their shared developer base.
Machine Learning & AI
One of the fastest-growing clusters, containing PyTorch, TensorFlow, Hugging Face libraries, and hundreds of model implementations and research codebases. The LLM sub-cluster has expanded dramatically in recent years and now forms its own distinct neighborhood within the broader ML region.
Rust Ecosystem
Despite being younger than most ecosystems on the map, Rust has formed a remarkably tight and coherent cluster. Systems programming, WebAssembly, embedded systems, and Rust web frameworks sit in close proximity — reflecting the community strong internal cross-project interest.
Data Visualization
D3.js, Observable, Plotly, and chart-building libraries form a cluster that bridges the JavaScript and data science regions — a visual reflection of data visualization position at the intersection of software engineering and data analysis.
Developer Tools & DevOps
Git tools, CI/CD systems, containerization (Docker, Kubernetes), and terminal utilities form a cross-language cluster near the center of the map — close to many other ecosystems because developers across all languages use these tools.
Frequently asked questions
- What is the Map of GitHub?
- An interactive visualization that places 700,000+ GitHub repositories in a 2D space based on co-star similarity — repositories frequently starred by the same developers are placed near each other.
- Who made the Map of GitHub?
- The Map of GitHub was created by developer Andrei Kashcha (anvaka on GitHub). This hosted version is based on his original open-source project and dataset.
- Is the Map of GitHub up to date?
- The map is rebuilt periodically. The current edition covers data through the most recent computation by anvaka pipeline. Individual repository growth between updates is not reflected in real time.
- Why isn't my repository on the map?
- Repositories need a minimum number of stars to appear on the map — this threshold filters out inactive forks and low-engagement projects to keep the map readable. If your repository is below this threshold it will not appear, but may be included in future editions as it grows.
- How is this different from GitHub's Explore page?
- GitHub Explore uses topic tags and trending metrics. The Map of GitHub uses co-star behavioral data — which projects developers actually use and star together — producing a more nuanced picture of ecosystem relationships.
- Can I find repositories in a specific programming language?
- Yes — zoom into the map and each major language cluster becomes visible. The JavaScript, Python, Rust, Go, and other language regions are distinct enough to navigate to by exploring the map.
Source code: github.com/anvaka/map-of-github ↗
Try City Roads next
Switch from stars to streets — minimalist OpenStreetMap linework for any city on Earth.
Open City Roads →