# GitHub Records 1.3 Billion Merged Pull Requests in 12 Months By Simon Yoon Canonical URL: https://www.tokenpost.com/news/technology/43090 Published: 2026-10-29T00:18:20.000Z Updated: 2026-10-29T00:18:20.000Z Section: Technology > The tally covers Oct. 1, 2025, through Sept. 30, 2026, and its agent measure can miss some agent use. GitHub logged 17.9 billion pushes, 34.3 billion commits and 1.3 billion merged pull requests in the year through Sept. 30, as detectable coding-agent drafts spread across the platform. The 12-month tally covers Oct. 1, 2025, through Sept. 30, 2026, and activity figures generally include public and private repositories. Merged pull requests averaged 3.5 million a day. The figures set new highs across all three measures. GitHub recalculated prior-year figures using the same definitions and period. The totals measure different actions. A push can contain multiple commits, while a merged pull request is counted separately. The annual figures add context to [earlier coverage of GitHub’s September commit surge](). More than 40% of September pull requests began with a detectable coding agent’s first draft, compared with 17% in March. GitHub counted 110 million such pull requests in September, about six times March’s volume. The measure relies on detectable agent signals, so it can miss some agent use. New repositories with an agent configuration file averaged eight times as many commits as those without one. The comparison does not show that configuration files caused higher activity; repositories without them may also use AI. In September, 90% of agent-involved pull requests merged within two weeks, compared with about 86% of pull requests without detected agent involvement. Those are observed rates and do not show that agent use caused a merge. A separate survey of open-source contributors and maintainers measured respondents’ experiences with review work. It had 234 qualified respondents, with 174 responses used for its review-time findings. Among maintainers at projects with more than 50 contributors, 64% said AI-assisted contributions lengthened review cycles; the reported figures were 55% at medium-size projects and 46% at small projects. Fielded from April through June, the survey measures respondents’ experiences, not activity across GitHub. Jay Parikh, Microsoft executive vice president of CoreAI, said, “Completed work, not more code, is the output that matters.” ## Links in this article - [earlier coverage of GitHub’s September commit surge](https://www.tokenpost.com/news/technology/29731)