People go looking for GEDmatch alternatives for two different reasons, and they need different answers. If the reason is DNA matching — finding relatives — that is a database question, and no analysis site substitutes for the testing companies' match lists. But if the reason is the admixture side — the calculators, the Oracle, the percentages — the alternatives are genuinely better than the original now, because GEDmatch's calculators froze around 2012–2016 and the ancient-DNA record they never saw has since grown past forty thousand published samples.
Here is what replaces what, all free, all in the browser.
Replacing the calculators: era-scoped Global25 fits#
The GEDmatch classics fit your kit against component constructs trained on the samples of 2012. The current equivalent fits a Global25 coordinate against curated panels of dated ancient populations, era by era — Late Bronze Age through Modern — with a stated fit distance: the Global25 admixture calculator, no account needed. Where K13's "West Asian" blends ancestries the field has since separated, an era panel names its sources and dates them.
The one setup step is the coordinate itself, which comes from the independent Eurogenes service — the full how-to is here, including the done-for-you route. Note the trade honestly: GEDmatch takes your raw file directly with no prerequisite, and its calculators, period pieces or not, remain readable instruments once you know their home fields.
Replacing the Oracle: distance rankings with dates#
Oracle's question — which reference populations sit closest to me? — outlived its implementation. The G25 distance tool answers it by plain Euclidean distance against 1,535 curated populations across six eras: the modern era for "who do I resemble today", the ancient eras for "whose excavated world do I land in". No component detour, no 2012 references, same habit of reading neighbourhoods rather than winners.
Replacing the scatter plots: PCA among published samples#
GEDmatch never had a good answer for seeing the space. The G25 PCA viewer projects your coordinate onto era reference views beside published ancient and modern samples, or fits a fresh PCA over rows you paste. It is the fastest way to catch the classic misreads — a "close" population that is actually alone in a sparse region, a mixture that averages you into empty space.
Replacing the utilities#
The odds and ends GEDmatch handled with 2012-era tooling all have current free equivalents here: raw file diagnostics (format, chip generation, usable markers per chromosome — before you upload anything anywhere), Y-DNA and mtDNA haplogroup finders, a coordinate averager, and a G25 authenticity check for rows of unknown origin. The full shelf is at the Lab hub.
What has no free replacement#
Honesty about the edges: GEDmatch's matching features (one-to-many, shared segments with other users) live on its user database and are not what analysis sites do, and Ancestrify does not offer a stranger-pool match list. And no free tool anywhere replaces a tested model: when a percentage needs to survive argument, the step up is qpAdm — explicit sources, stated outgroups, a p-value that can reject — which is a different kind of statement, built by hand, and priced accordingly.
For everything the admixture side of GEDmatch actually did, though: the current references are better, the tools are faster, and the price is the same zero.
Terms used here are defined in the glossary.
