Can AI build a family tree app like Ancestry?

Build a personal, read-only family tree viewer web app for my own use, to view a GEDCOM file I already exported for free from Ancestry. This app does not replace Ancestry's records, DNA matching, or Hints. It only displays a tree I already own.

Tech rules:
- Plain HTML, CSS, and vanilla JavaScript in one file. No framework, build step, account, or server.
- Store the imported GEDCOM data in localStorage under one key, with a version number.

Data model:
- Person: id, givenName, surname, birthDate (exact | approximate | unknown), deathDate (exact | approximate | unknown | living), sex, notes.
- Relationship: id, personId, relatedPersonId, type (parent | spouse | child | sibling | half-sibling | adoptive-parent | adoptive-child), marriageOrder (for multiple marriages).

Features:
- Import a GEDCOM file and parse it into the Person and Relationship data model above.
- Display a simple tree or list view: pick a person, see their parents, siblings, spouses, and children.
- Search by name across the imported tree.
- A "living" flag on any person marked alive that hides their birth date and notes from the default view, for privacy.
- Export whatever notes or edits I've added locally back out as JSON, separate from the original GEDCOM.

Edge cases the build must handle correctly:
1. Multiple marriages and step-relationships must display without merging into a single ambiguous spouse relationship.
2. Half-siblings and full siblings must be visually distinguishable, not lumped into one "sibling" category.
3. Adoptive and non-biological parent-child links must be labeled distinctly from biological ones, not hidden.
4. Unknown, partial, or approximate dates (for example "circa 1850") must display as entered, never rounded or guessed into a fake exact date.
5. The same name repeating across generations must not cause two different people to be merged into one record.
6. Two branches that merge at a shared ancestor must reference the same person record, not create a duplicate.
7. Living relatives marked as such must have their details hidden by default, while deceased relatives show full detail.
8. Non-traditional family structures need flexible parent slots, not a fixed "father and mother" pair per person.

Backup:
- One-click "Export JSON" for any local notes or edits. "Import JSON" restores after confirmation.
- GEDCOM import as described above, read-only against the original file structure.

Layout:
- Mobile-first, readable at 375px wide.

Exclusions:
- No historical records search, no DNA matching, no Hints, no public tree sharing, no publishing, App Store submission, accounts, payments, or social features.

Write a short README explaining where the data is stored, how to export local notes, and that this app is a personal viewer for a tree you already built on Ancestry for free, not a replacement for Ancestry's records or DNA features.

Keep it under 500 lines. This is a personal tool, not a product.

What is the short answer?

No, and there’s not really anything to build here. Ancestry’s tree-drawing feature, the part most people imagine paying for, is already free and unlimited on a Guest account. Full GEDCOM export is free too, on any tier.

The paid membership, U.S. Discovery starting around $19.99/mo on a 12-month term [VERIFY], buys something else entirely: billions of licensed historical records, AncestryDNA matching, and the Ancestry Hints network that surfaces matches by cross-referencing your tree against millions of other members’ trees. None of that is buildable by one person.

Option	Best for	Upfront cost	Ongoing cost	Main limitation
Build a rough prototype	Developers curious about parsing GEDCOM files as a learning exercise	$0 and 2–3 hours	$0	Duplicates a free feature Ancestry already gives you
Build a daily-use personal version	Nobody, really. Use Ancestry’s free account for the tree instead	$0	$0	The build adds nothing the free tier doesn’t already do
Keep paying for Ancestry	Anyone who wants historical records, DNA matching, or the Hints network	$0	$19.99–44.99/mo depending on tier [VERIFY]	Price buys records and matching, not the tree itself

If you still want a local, offline viewer for a tree you’ve already exported, the starter prompt at the end of this page builds that, nothing more.

What should the first version include?

Honestly, the first version should be your free Ancestry account. If you want a personal, offline copy of your tree anyway, for privacy or portability, keep it narrow:

• Import a GEDCOM file you’ve already exported for free
• Basic person and relationship display: parents, siblings, spouses, children
• A “living” flag that hides details for people still alive
• Deliberate exclusion: no records search, no DNA matching, no Hints. None of that data exists outside Ancestry, so a build can’t offer it.

What can AI build reliably?

AI can parse a GEDCOM file and build a simple viewer: search by name, click a person, see their immediate relationships. That’s a modest, well-scoped task.

AI cannot build Ancestry’s actual paid product. It has no access to billions of digitized historical records from archives, no lab-processed DNA database of tens of millions of testers, and no network of millions of other users’ trees to cross-reference against for Hints. Those aren’t coding problems. They’re licensing, lab infrastructure, and network effects that took a company years to build.

Where will AI need human help?

If you do build the personal viewer, the tricky parts are all genealogy data-modeling problems, not technical ones:

• Multiple marriages and step-relationships. Decide how you want these displayed before building, since the data model needs to support more than one spouse per person over time.
• Half-siblings versus full siblings. These need to be visually distinct, or the tree becomes misleading.
• Adoption and non-biological links. Label these clearly rather than hiding them inside a generic “parent” relationship.
• Unknown or approximate dates. Never let the app guess or round a “circa 1850” into a fake exact year.
• Duplicate people across generations. The same name repeating needs careful handling so two different ancestors don’t merge into one record, and so two branches that share an ancestor correctly point to the same person.

None of this replaces what Ancestry’s subscription buys. It’s just careful data entry for a tree you already own.

Can I get my data out of Ancestry?

Yes, and it’s free on any tier, including the free plan. The tree owner exports a full GEDCOM 5.5.1 file from Tree Settings > Manage Your Tree > Export Tree.

The trap: GEDCOM excludes photos, documents, and other uploaded media, which you have to download separately, and it excludes DNA match data entirely. The export is a static snapshot, not a live sync. Some sources report that unusual characters and detailed note formatting don’t always transfer cleanly into other genealogy software [VERIFY].

The starter prompt’s viewer imports this same GEDCOM file. Since export is already free, this isn’t solving an access problem, just giving you an offline copy if you want one.

How long will it take and what will it cost?

A rough prototype of the GEDCOM viewer: 2–3 hours on a free AI plan. $0.

A daily-use version with search and relationship display: add another 1–2 hours. $0.

Replacing Ancestry’s actual paid product: not possible at any budget or build time. The records, DNA data, and Hints network aren’t things one person can assemble.

Ancestry’s U.S. Discovery tier runs around $19.99 per month on a 12-month term [VERIFY]. Building the viewer doesn’t save you this cost if you still want records or DNA matching. It only helps if you were never going to use the paid features anyway.

These are estimates from reviewing the task, not measurements from a completed build.

Which AI tool or approach should I use?

Use ChatGPT or Claude if you want the GEDCOM viewer built from the starter prompt. Either works on a free plan for a single-file app like this.

Don’t reach for Lovable or Cursor expecting to build anything that touches actual historical records or DNA data. There’s no dataset for either tool to work with, since Ancestry’s records are licensed and DNA data is lab-processed, not something an AI coding tool has access to.

What will I need to maintain?

• Re-export a fresh GEDCOM file from Ancestry whenever you add people there, and re-import it into the viewer
• Nothing else, since the viewer has no background process or sync

Should I build it, buy it, or reduce scope?

Build it only if you want an offline, private copy of a tree you’ve already built for free on Ancestry, and you’re fine re-importing GEDCOM manually each time it changes.

Buy it if you want historical records, AncestryDNA matching, or the Hints network. At $19.99–44.99 per month depending on tier [VERIFY], that’s what you’re actually paying for, and none of it is replaceable by a personal build.

Reduce scope by simply using Ancestry’s free Guest account for the tree itself. This is the honest answer for most people who search for a “family tree app” wanting to sketch relatives without paying: the free tier already does the buildable part.

Recommendation: don’t build a replacement. Use Ancestry’s free tree builder directly. Only add a personal GEDCOM viewer on top if you specifically want an offline copy, and only pay for a membership if you actually want records or DNA matching.

What if I wanted to ship this to other people?

A GEDCOM viewer could be shared as a simple open tool for other genealogy hobbyists to view their own exported files. It would need to stay strictly a viewer, since anything touching real historical records or DNA data would require the same licensing and lab infrastructure Ancestry has, which a shipped side project can’t replicate.

Starter build specification

• User: You, viewing a family tree you’ve already built for free on Ancestry.
• Problem: Wanting an offline, private copy of your tree, not paying for records or DNA matching you don’t need.
• Core workflow: Export GEDCOM from Ancestry for free, import it into the viewer, browse relationships locally.
• Required features: GEDCOM import, person and relationship display, search, a living-relative privacy flag.
• Deliberate exclusions: No records search, no DNA matching, no Hints, no public tree sharing, no store submission.
• Data ownership: GEDCOM data and any local notes stored in localStorage. Notes exported as JSON.
• Definition of done: You import your GEDCOM file, search for a relative by name, and see their parents, siblings, spouses, and children correctly, including any half-siblings or approximate dates from the original file.

For the principles behind which apps are safe to build yourself, read can AI build an app for me. If you’re organizing family photos or memories alongside a tree, see the Day One alternative. Other well-scoped personal trackers on this site, like a habit tracker or a pet record keeper, are better examples of what a small build should replace than a family tree app is.

Sources and verification

• Ancestry subscription plans explained (Family Tree Magazine): tier breakdown and what each includes.
• Ancestry’s best membership guide (Family History Daily): pricing detail across U.S. Discovery, World Explorer, and All Access.
• Can I export my family tree as a GEDCOM file (Ancestry Support): GEDCOM export steps and what’s included, free on any tier.

This app was not built and tested for this page. Ancestry.com and support.ancestry.com returned errors on direct fetch during research, so pricing and export mechanics are sourced from cross-checked third-party genealogy publications rather than Ancestry’s own pages directly, and annual totals varied slightly between sources [VERIFY]. Ancestry’s price was checked on September 24, 2026.

Reviewed on September 24, 2026.