Count Isn't Weight: What 67,000 Repos Say About Language Ecosystems

The usual way to compare language ecosystems is to ask which language has the most projects. It's an easy question to answer and a weak one to act on — it tells you about count, not about weight.

We track more than 67,000 high-star GitHub repositories and record how their stars move every day. Splitting that corpus by language shows something more useful than a headcount.

The raw numbers

This table is read live from our database on every page load — it isn't a screenshot:

Three things worth noticing

1. The leader by count doesn't lead by median

Python is far ahead on repository count — roughly 1.4× the runner-up. But look at the median star count, which answers "if I pick a random project in this language, how popular is it likely to be": Python sits behind both Go and TypeScript.

That's a signal of breadth. Python covers scientific computing, scraping, machine learning, automation — almost everything. The flip side of breadth is a very long tail, with many projects sitting just above the tracking threshold.

2. Go has the highest hit rate

Go has about a third of Python's repository count, yet the highest median of any language here.

The plausible explanation is domain concentration. Go's popular projects cluster into cloud-native infrastructure, DevOps and CLI tooling — fields whose users are themselves developers, and therefore people who star things. Much of Python's reach is into audiences that never open GitHub at all.

3. Total stars and median tell different stories

JavaScript's total star count is high while its median is comparatively low. The total reflects a cohort of very large, very old projects — that's accumulated history. The soft median suggests a smaller share of recent projects break through.

TypeScript inverts this: fewer repositories than JavaScript but a higher median, consistent with newer projects defaulting to TypeScript from day one.

The top of the list

Composition matters as much as counts. These are the highest-starred repositories we track right now:

Notice how many of them are learning resources and curated lists rather than runnable software. Those earn enormous star counts while saying little about a language's engineering ecosystem — lumping them in with frameworks and runtimes distorts any comparison.

What this data can't tell you

Any analysis should state its own limits:

• The sample is already-popular projects, not all of GitHub. The medians here are medians among projects that already cleared the bar, far above the true site-wide figure.
• Stars aren't quality, and they aren't usage. They approximate developer visibility. Plenty of critical enterprise libraries carry star counts well below their real importance.
• Language attribution follows GitHub's primary-language detection. A frontend framework can be labelled something unexpected because of how its file bytes shake out.
• Ecosystems differ in age. C and Java have decades of accumulation; Rust has had far less time. Comparing totals directly is unfair to the younger ones.

Takeaway

If you're choosing a stack, a better question than "which language has more projects" is: how many mature options exist in the specific niche I need. Global counts help surprisingly little with that.

The table above updates itself with our daily crawl. You can slice the same data by language and category yourself on the rankings.