GitHub Global Stats

Is AI changing how we code? Monthly repositories, issues, and pull requests created across GitHub since 2008.

The numbers behind the charts

GitHub activity by year

PeriodNew repositoriesIssues openedPull requestsCommits *
2026 (in progress)57,996,80124,321,404100,159,7871,374,697,048
202561,769,67119,912,06982,640,4051,511,364,038
202444,535,95215,476,91457,262,4141,065,905,888
202338,966,35515,465,76456,793,4281,034,417,715
202234,089,83313,292,79977,330,1681,003,968,359
202128,150,25612,800,33161,624,734989,911,139
202021,575,86112,195,23441,960,3421,099,172,965
201915,826,0919,837,29021,043,0811,054,969,776
201813,337,4498,866,75313,266,3471,186,193,648
201710,683,8247,627,6409,937,2731,440,035,742
20167,684,73111,134,0007,659,7661,622,075,276
20155,143,44813,070,1364,713,9951,705,399,915
20143,187,5253,326,5102,662,4371,838,048,678
20131,859,3272,156,5621,445,8321,775,680,390
2012971,2871,105,696635,8441,716,247,206
2011428,452589,251240,4561,569,678,170
2010181,059228,53631,9111,386,925,867
200978,018107,72401,403,330,087
200820,53052,1530988,711,089

Last 12 months

PeriodNew repositoriesIssues openedPull requestsCommits *
August 2026 (in progress)3,222,5101,769,6866,993,279
July 20268,896,1464,542,09517,914,471186,274,257
June 20268,829,8643,718,27615,484,076225,363,186
May 20268,469,2083,645,41714,473,713238,866,993
April 20268,007,8223,246,63212,831,810210,450,230
March 20267,947,0993,079,96912,669,170209,804,884
February 20266,409,4102,337,55210,257,431155,722,979
January 20266,214,7421,981,7779,535,837148,214,519
December 20255,722,3311,716,3018,697,233139,500,175
November 20256,030,8881,816,8688,721,459144,863,423
October 20255,862,2832,155,2158,822,517139,353,409
September 20255,673,8681,802,1998,020,760134,888,559

* Commit counts use git metadata dates and are inflated in earlier years by imported repositories — see the FAQ below. Periods marked "in progress" are partial.

FAQ

Is AI actually changing how we code?

That's the question this page exists to explore. Every chart is annotated with the release dates of major AI models and coding tools — GitHub Copilot, ChatGPT, GPT-4, Claude, DeepSeek and others — shown as dashed vertical lines, so you can look for inflection points in global GitHub activity around those moments. The "AI" switch on each chart toggles the milestone labels.

Where does the data come from?

Directly from GitHub. Each data point counts how many repositories, issues, or pull requests were created across all of GitHub in a given month — or on a given day, if you click into a month for the daily drill-down.

How fresh is the data?

The data is refetched every day, and this page is rebuilt right after each fetch. The data was last fetched on August 12, 2026.

Why the asterisk on commits?

Commit counts rely on git metadata dates rather than the date the code actually reached GitHub. Repositories imported with long pre-existing histories — the Linux kernel, old SVN migrations — inflate earlier years, which makes commits unreliable for measuring GitHub activity over time. Repositories, issues, and pull requests don't have this problem: they're counted by their GitHub creation date.