[2/3] Octoverse 2025: A New Developer Joins GitHub Every Second as AI Leads TypeScript to #1 (Category Engineering)
Continuing my discussion of GitHub’s annual report on its software-development ecosystem, let’s look more closely at the results.
📈 Unprecedented community growth
- More than 36 million new developers joined the platform, the largest absolute increase on record.
- Developers created around 121 million new repositories during the year, bringing the total to 630 million.
- Commits reached almost 1 billion in 2025, up 25% from 2024.
- Around 43.2 million pull requests were merged each month, up 23% from 2024.
This growth coincided with the launch of free GitHub Copilot at the end of 2024, which attracted a wave of new users and projects.
🤖 AI has become integral to development Over the year, generative AI moved from novelty to a standard developer tool.
- GitHub now has around 4.3 million AI-related repositories. This estimate uses repository tags and classifies some tags as associated with AI projects.
- More than 1.1 million public projects use SDKs for large language models, and 693 thousand of them were created in the past 12 months: 178% growth year over year.
- Newcomers adopt AI immediately: almost 80% of new developers start using GitHub Copilot autocomplete within their first week of coding.
AI tools are now an expected part of the developer environment. Growing adoption accompanies increasing code contributions: a record 518.7 million pull requests were merged during the year, up 29%, while contributions to open projects exceeded 1.12 billion, up 13% year over year.
🚀 TypeScript takes the #1 spot among languages For the first time in a decade, GitHub’s language leader changed: in August 2025, TypeScript overtook Python, as well as longtime leader JavaScript, to become the platform’s most-used language. Two factors explain its rapid growth:
- Modern frontend frameworks, including React and Angular, generate TypeScript project templates by default.
- Teams increasingly prefer strongly typed code when using AI. TypeScript’s extensive type system makes AI-generated code more reliable.
TypeScript’s community grew by approximately 1.05 million developers over the year, up 66%, while Python added around 850 thousand, up 48%. Despite losing first place, Python remains central to AI and data science, with 2.6 million active contributors. Jupyter Notebook remains a main environment for experimentation, appearing in approximately 403 thousand repositories within AI projects. Together, the TypeScript and Python communities now include more than 5.2 million developers, around 3% of all active GitHub users.
This shift supports the view that AI affects technology choices as well as coding speed: teams are more willing to entrust AI-generated code to strongly typed languages, particularly for production systems.
💯 AI is changing development preferences and practices Octoverse points to a clear connection between AI-tool adoption and the evolution of developers’ technology stacks:
- From 2025, Python’s growth runs almost parallel to JavaScript/TypeScript, suggesting AI’s broad influence across ecosystems, from web development to data analysis.
- New ways of working with code are emerging, including vibe coding.
- AI assistants have lowered the entry barrier considerably. If AI continues simplifying software creation, the number of people able to code could rise sharply.
- An era of autonomous AI agents is on the horizon. Tools already exist that can generate code, run tests and even open a pull request for a person to review.
In the concluding post, we’ll discuss open source, AI’s effect on development productivity and the report’s conclusions.
#AI #ML #Software #Architecture #Processes #DevEx #Devops #Metrics #Engineering #OpenSource #Architecture