Be one with code

Grokbit is the best way to search, browse, and explore code.

Enter any public GitHub repo and try it out:

We currently parse JavaScript, C, C++, Java, C#, Ruby, and Python. With your help, we hope to parse many more languages and build the ultimate open source tool for code search and analysis.

Our goals

  • We want to give you blazing-fast code-aware search. Immediately find where a class is defined, a function called, or a property documented. Get real-time suggestions as you search, even across huge repos. Use punctuation like + or => in your queries and get the right hits.

  • We want to deliver features like automatic refactoring, function call graphs and code completion to our beloved open source editors. Let's build code intelligence into a shared open source tool, then plug it into various editors.

  • We want to enable plugins that can give you a meaningful overview of a repo, find bugs, and do other semantic tasks. An easy way to feed code as features into machine learning models wouldn't hurt, either.


Semantic search

Our search engine understands code. For example, we sort definitions first because we know that's often what you're looking for. So if you search for object in Git or parseHTML in jQuery, we hit the spot immediately. We understand the semantics of each hit.

Linux Semantic Search

Advanced operators

We have code-specific operators like type, fn, or param. You can specify programming language via lang and fine-tune the search via path. Use * with any operator to make searching a breeze. Operators can be combined and mostly do what you'd expect.

Searching with fn: and glob

See the forest for the trees

It's hard to dig into an unknown codebase. We currently provide basic stats to show you where the code is, languages used, and so on. We plan to give many more insights into a repo.

Code-aware navigation

Click on identifiers to see their definition, declaration, and so on.


We are standing on the shoulders of giants like ctags and OpenGrok. But we want to go further: more sophisticated parsers, complete knowledge of the codebases (all tokens, not just tags), and ridiculous performance (right now, we can index the Linux kernel in about 45 seconds). Ultimately, we hope this will improve some aspects of the programming experience.

Pulling this off requires a wide mix of technologies. We've built lightning-fast parsers in the backend along with innovative data structures to power search and navigation. In the frontend we use web workers, websockets, and dynamic rendering to deliver the best possible UX.

Let's talk If this sounds interesting, maybe you should join the project. We plan to have repos open on GitHub soon, but if you'd like to get started, drop us a line.


Soon we'll be releasing source code and more features. Sign up to hear about it.

Try Grokbit

Here are suggestions for repos you can try out. But remember, you can take any GitHub URL and simply replace "github" with "grokbit" to load the code in our interface. This works even for branches and tags, like this. We have parsers for JavaScript, C, C++, Java, C#, Ruby, and Python. Other languages appear as plain text for now.