On Oct 24, 2026 the median repo we track gained 1 star in a day. Of the 12,464 repos with a gain for that day, 3,587 gained none, 1,560 lost stars, 262 gained 100 or more and 10 gained 1,000 or more, so a good day depends mostly on how big the repo already is.
This guide is for a developer or maintainer who sees “+40 stars today” next to a repo and wants to know whether that is a lot. The figures are RepoTrending’s own count of one day, taken from the star counts we save for each repo every day. The repos are the ones our searches find: every public repo with 5,000 stars or more that they reach, plus recent ones, so the numbers describe popular repos, not GitHub as a whole.
How many stars does a typical GitHub repo gain in a day?
One, at the median, and none for more than a quarter of repos. We counted the stars each repo gained in the 24 hours to its Oct 24 snapshot, which is the same count the site uses for its “today” ranking.
- 1star gainedthe median repo
- 1,560repos lost stars
- 262repos gained 100 or more
- 10repos gained 1,000 or more
| Stars gained on Oct 24 | Repos | Share of 12,464 |
|---|---|---|
| Lost stars | 1,560 | 12.5% |
| None | 3,587 | 28.8% |
| 1 to 9 | 5,359 | 43.0% |
| 10 to 99 | 1,696 | 13.6% |
| 100 to 999 | 252 | 2.0% |
| 1,000 or more | 10 | 0.1% |
Only 1,958 repos, 15.7% of those we track, gained 10 or more stars. The biggest single gain was 6,448 stars, for morluto/rea, which had 373,156 stars at the end of the day.
A repo can also go down. Someone who starred a repo can remove their star, so 1,560 of the repos ended the day with fewer stars than they started it with. 1,465 of them lost one or two stars. The biggest drop was tek-sys-hub/Rubix-lang, from 1,902 stars to 29. We don’t know why, and one drop is not proof of anything.
Does the size of a repo change what a good day is?
Yes, a lot. The bigger the repo, the more stars it gains on a median day, which fits with more people seeing it, though we did not measure that.
| Stars in total | Repos | Median gain | One repo in ten gained more than | Repos with 100 or more |
|---|---|---|---|---|
| Under 5,000 | 500 | 14 | 106 | 54 |
| 5,000 to 9,999 | 6,586 | 0 | 6 | 28 |
| 10,000 to 19,999 | 3,145 | 1 | 12 | 33 |
| 20,000 to 49,999 | 1,725 | 4 | 34 | 51 |
| 50,000 to 99,999 | 373 | 16 | 108 | 43 |
| 100,000 or more | 135 | 75 | 373 | 53 |
A repo with 100,000 stars or more gains 75 on a median day and 373 on a day in its top tenth. A repo with 5,000 to 9,999 stars gains none at the median, and 6 on a day in its top tenth. The same +40 is a quiet day for the first and an unusual day for the second.
The first row is not a fair sample. The 500 repos under 5,000 stars are recent repos our searches found because they grew, so their median of 14 says what those repos did, not what small repos do.
What gain puts a repo near the top of the day?
About 50 stars puts a repo among the top 4%, and 100 puts it among the top 2.1%. We counted how many of the 12,464 repos reached each level.
| Gain of at least | Repos | Share of 12,464 |
|---|---|---|
| 10 | 1,958 | 15.7% |
| 50 | 501 | 4.0% |
| 100 | 262 | 2.1% |
| 500 | 38 | 0.3% |
| 1,000 | 10 | 0.1% |
These cutoffs are one day’s. The ranking on the home page puts the repos that have a page in order by exactly this gain, and daily, weekly or monthly explains why a repo that does well one day may not on the next.
How do you read your own repo’s number?
Find its group, compare it with repos of its size, and don’t decide on one day. If you maintain a repo with 15,000 stars that gained 12, you were in the top tenth of repos your size that day. If a repo with 150,000 stars gained 12, it had a quiet one.
A gain says how many stars moved, not why. A launch post, a mention in a newsletter or a change of the README can each move it, and so can stars we can’t tell apart from real ones. For what to check when a gain is much bigger than the group it belongs to, read how to read a big one-day star gain. For why we rank by a gain and not by the total, read stars gained vs total stars.
What does this sample hide?
It hides every day but one, and every repo we don’t find. Our searches reach repos with 5,000 stars or more and recent repos; how GitHub Trending works and how we rank explain the difference between ours and GitHub’s. A repo that is small, quiet and takes off is not found until a search reaches it.
We don’t check for bought or farmed stars yet, so a gain here is what GitHub shows and nothing more. The same goes for a drop. The first weekly gains come on Oct 30 and the first monthly ones on Nov 22; when they do, this guide will be recounted.
Frequently asked questions
How many stars does a GitHub repo gain in a day?
Few. On Oct 24, 2026 the median of the 12,464 repos we track gained 1 star, 3,587 gained none and 1,560 lost stars. Only 262 gained 100 or more.
How many stars a day makes a repo trend?
It depends on the size of the repo. On Oct 24, 501 of 12,464 repos gained 50 or more stars, and 262 gained 100 or more. The median repo with 100,000 stars or more gained 75.
Can a GitHub repo lose stars?
Yes. A star can be removed by the person who gave it, so a count can go down. 1,560 of the 12,464 repos we track ended Oct 24 with fewer stars than a day before.
Is one day of stars enough to judge a repo?
No. One day shows the size of a gain, not whether it lasts. Our first weekly gains arrive Oct 30 and the first monthly ones Nov 22.
Where to look next
- The home page: today’s ranking, in order of the stars gained.
- The all-time list: the most starred repos, for the totals behind the gains.
- How we rank: how the repos are found and how a gain is counted.
Data used
Source for every row: RepoTrending, GitHub REST API (public repository search, read-only), snapshots of Oct 23 and Oct 24, 2026 (collector export gains.json, run of 2026-10-24), counted by RepoTrending. A gain is the star count at the Oct 24 snapshot minus the count 24 hours earlier, read the way collector/compute.mjs reads it. The 32 repos first seen on Oct 24 have no gain yet and are left out. Medians of an even count are the mean of the two middle values; “one repo in ten gained more than” is the value at position floor(0.9 times (n minus 1)) of the sorted gains of the group.
| Number | Source | Date |
|---|---|---|
| 12,464 tracked repos with a gain for the day (12,496 tracked, 32 first seen on Oct 24) | RepoTrending meta.json and gains.json |
Oct 24, 2026 |
| Median gain 1; no change 3,587; lost stars 1,560; 1 to 9 stars 5,359; 10 to 99 1,696; 100 to 999 252; 1,000 or more 10 | RepoTrending gains.json |
Oct 24, 2026 |
| Gain of at least 10: 1,958; at least 50: 501; at least 100: 262; at least 500: 38; at least 1,000: 10 | RepoTrending gains.json |
Oct 24, 2026 |
| By total stars (repos, median gain, tenth, repos with 100 or more): under 5,000 500, 14, 106, 54; 5,000 to 9,999 6,586, 0, 6, 28; 10,000 to 19,999 3,145, 1, 12, 33; 20,000 to 49,999 1,725, 4, 34, 51; 50,000 to 99,999 373, 16, 108, 43; 100,000 or more 135, 75, 373, 53 | RepoTrending gains.json |
Oct 24, 2026 |
| morluto/rea: gained 6,448, 373,156 stars in total | RepoTrending gains.json |
Oct 24, 2026 |
| Repos that lost one or two stars: 1,465 of the 1,560 | RepoTrending gains.json |
Oct 24, 2026 |
| tek-sys-hub/Rubix-lang: 29 stars at the end of the day, a gain of minus 1,873, so 1,902 the day before | RepoTrending gains.json |
Oct 24, 2026 |
| First weekly gains Oct 30; first monthly gains Nov 22 | RepoTrending meta.json (windows) |
Oct 24, 2026 |
More guides
- How to check a GitHub repo before you depend on it
- How to read a big one-day star gain on GitHub
- Stars per year: comparing old and new GitHub repos
- Daily, weekly or monthly: which window to watch?
- Find Claude Code repos worth trying (Oct 2026)
- GitHub trending alternatives: what each one counts
- How does GitHub decide what is trending?
- Open source licenses on popular GitHub repos
- Stars gained vs total stars: which says more?
- Trending badge for your README, and what it says
- What 12,468 GitHub repos look like