Google Maps Timeline

Your Google Maps Timeline knows every country you have been to

Written 23 September 2026  ·  by Aki Takashima, Shiori Studio

If Location History was on, Google Maps has been keeping the list you never wrote down: every country you have been to, and when. The catch is that you can no longer open it on a computer. Google's own help page now puts it plainly: “Timeline is not available for Maps on your computer.”

The history still exists. It lives on your phone, and the phone will hand it over as one JSON file. This page is how to get that file and turn it into an actual answer — a count, a list with years, and a map — without sending a record of everywhere you have been to anyone's server.

Where Timeline went

Timeline used to be a page on the Google Maps website. Google moved it onto the phone: according to its Timeline help page, visits and routes are now saved to the Timeline on each signed-in device, and a backup, if you turn it on, is an encrypted copy stored on Google's servers.

Two consequences matter here. First, there is no longer a map of everything on a big screen, which is why people end up searching for a “Timeline viewer”. Second, what is on your phone is what you have. If older years are missing from it, an old Google Takeout archive, if you ever downloaded one, may still have them. That case is further down.

Export it from your phone

These are the paths from Google's help pages as of September 2026. Menus move around; if a name has changed, the export button is still somewhere under Timeline or Location settings.

To get the file onto a computer, AirDrop or a cable is the tidy way. Think twice about emailing it to yourself or dropping it into a shared drive: it is a complete record of where you sleep, work and travel.

Turn it into a list of countries

Open driftlog-counter.pages.dev and drop the file on it. On a phone you do not need a computer at all: on an iPhone, tap Choose photos → Choose Files and pick location-history.json from where you saved it. What happens next:

• It reads the file in slices of a few megabytes at a time, so a Timeline with years of history does not freeze the tab. Visits, routes and raw location signals are all used.
• Every point is tested against country borders — Natural Earth at 1:50m, with the smallest countries at 1:10m — rather than matched to the nearest town, so points near a border land on the right side and Singapore, Monaco or Vatican City are not silently merged into a neighbour.
• You get a count, a list and a map. Each country shows the first and last year it appears in your history. The headline counts the 193 UN member states plus Taiwan; territories such as Hong Kong or Puerto Rico are listed separately and not counted. The map can be saved as an image, or on a phone sent straight to Photos or a message with Share.
• Nothing is uploaded. The page makes no requests with your data, has no analytics and sets no cookies. Load it, switch off Wi-Fi, then drop the file in: it still works.

You can drop photos in at the same time — the page reads the location stored in photo files too — and the countries from both are merged into one list.

If you have an old Google Takeout archive

Before Timeline moved to the phone, Google Takeout could export Location History from the web. If you downloaded one of those archives, unzip it and look inside the Location History folder:

• Records.json holds every raw location point, often for many years. It can be hundreds of megabytes. In archives from before 2022 the same file is called Location History.json.
• Semantic Location History holds one file per month, with the places Google decided you visited.

Drop the whole unzipped folder on the counter, or just that one file. Both are read, and a large file is fine, because it is never loaded in one piece. If that archive is older than your current phone, it may cover years your on-device Timeline no longer has.

The same thing in Python, if you would rather see the code

All three formats bury coordinates in text in one of three shapes, so a regular expression gets every point out without caring about the JSON structure, which Google has changed several times:

import csv, re, sys

# Records.json ("latitudeE7"), Timeline.json from Android ("48.8584°, 2.2945°")
# and location-history.json from the iPhone ("geo:48.8584,2.2945").
POINT = re.compile(
    r'"latitudeE7"\s*:\s*(-?\d+)\s*,\s*"longitudeE7"\s*:\s*(-?\d+)'
    r'|(-?\d{1,2}\.\d+)(?:°|\\u00b0),\s*(-?\d{1,3}\.\d+)(?:°|\\u00b0)'
    r'|geo:(-?\d{1,2}\.\d+),(-?\d{1,3}\.\d+)'
)

out = csv.writer(sys.stdout)
for path in sys.argv[1:]:
    text = open(path, encoding="utf-8").read()
    for m in POINT.finditer(text):
        if m.group(1):
            lat, lon = int(m.group(1)), int(m.group(2))
            if lat > 900000000: lat -= 2**32      # a few old exports overflowed
            if lon > 1800000000: lon -= 2**32
            lat, lon = lat / 1e7, lon / 1e7
        elif m.group(3):
            lat, lon = float(m.group(3)), float(m.group(4))
        else:
            lat, lon = float(m.group(5)), float(m.group(6))
        out.writerow([lat, lon])

python3 timeline_points.py Timeline.json > coords.csv

That gives you a CSV of latitude and longitude, with no dates. For turning it into countries offline, and for why the obvious nearest-town lookup gets borders wrong, see Step 3 on the photo version of this page; that script expects a date in a third column, so with this CSV it counts points rather than days. This reads the whole file into memory, which is fine for a phone export and slow for a very large Records.json.

Where the count goes wrong