For more than a decade, the language at the top of GitHub did not change. That streak just ended. According to GitHub's 2025 Octoverse report, TypeScript passed both Python and JavaScript in August 2025 to become the most-used language on the platform by contributor count. It is the biggest ranking shift the report has recorded in ten-plus years, and the interesting part is not the new name on the board. It is the reason behind it.

The number that ended Python's run

TypeScript closed 2025 with more than 2.6 million active contributors, having added over a million in a single year. That is a 66% jump. The crossover itself was tight: TypeScript edged past Python by roughly 42,000 contributors. Python was not collapsing while this happened. It grew 48% and added about 850,000 new contributors of its own. JavaScript, the language TypeScript is built on top of, grew a comparatively sleepy 25%.

All of this played out on a platform adding developers faster than ever. GitHub now hosts more than 180 million developers and signed up 36.2 million of them in 2025, which works out to roughly one new account every second. At that scale, 43.2 million pull requests merge every month (up 23% year over year) and public repos took in 1.12 billion contributions. When the base is that large, even a small shift in what each new developer reaches for moves the aggregate rankings quickly.

AI is not in the background of this story

GitHub does not hedge on the cause. Its own framing says AI now shapes not only how fast code gets written but which languages developers pick in the first place. The supporting numbers back that up. Repositories importing LLM SDKs climbed 178% year over year to more than 1.1 million. Nearly 80% of new GitHub users turned on Copilot within their first week. The people pouring into the platform are arriving with an AI assistant switched on, and they are writing a lot of TypeScript.

Here is the mechanism I find most convincing. A 2025 academic study cited by GitHub found that 94% of compilation errors produced by LLMs were type-check failures. Read that again. The single most common way AI-generated code breaks is exactly the failure mode a type checker is built to catch before anything runs. TypeScript's compiler sits right on top of the mistakes these models make most often. GitHub put it plainly, pointing to a move toward typed languages that make agent-assisted coding more reliable in production.

So the rise is not fashion. When you generate code in volume, static types act as a cheap, automatic guardrail. They reject a whole category of bugs before review, before runtime, before they ever cost anyone an afternoon. That is a real return for almost no extra effort, and teams shipping AI-written code are noticing.

The framework defaults quietly settled it

There is a second force compounding the first, and it is less dramatic but just as decisive. The path of least resistance now starts typed. Next.js, Astro, SvelteKit, and Angular all scaffold new projects in TypeScript by default. A developer spinning up something fresh is writing types whether or not they made a deliberate decision to. Multiply that default across millions of new projects and you get exactly the curve Octoverse recorded.

What I'd actually do with this

If you are starting new web or full-stack work in 2026, default to TypeScript and stop treating it as the optional grown-up version of JavaScript. The tooling, the frameworks, and the AI assistants are all tuned for it now, and swimming against that current just costs you velocity.

If your team ships AI-generated code, lean on types as a safety net rather than a style preference. The compiler catches the precise errors these models produce most. Few quality controls are this cheap to adopt.

And do not read this as a signal to abandon Python. Python still leads in AI-tagged repositories, and it remains the language of data science, ML, scripting, and a great deal of backend work. The honest read is not a winner-take-all fight. The two are settling into complementary roles: a typed front-end and edge layer paired with Python where it already dominates. Migrating away from Python because of a leaderboard would be a mistake.

The part worth sitting with is the second-order signal. The headline says TypeScript versus Python. The actual news is that AI tooling has become a primary driver of which technologies developers adopt at all. The compiler that catches an LLM's favorite mistake won this round. Expect the same logic to reshape testing frameworks, build tools, and runtimes next, and watch which ones make agent-written code safer by default.

Sources

  • https://github.blog/news-insights/octoverse/octoverse-a-new-developer-joins-github-every-second-as-ai-leads-typescript-to-1/
  • https://www.infoworld.com/article/4080454/typescript-rises-to-the-top-on-github.html
  • https://www.mgsoftware.nl/en/blog/typescript-overtakes-python-as-most-used-language-on-github