WEIR
Get a key
Theme

Twitter Follower Count History: See Past Counts for Free

Find a Twitter follower count history for free in Wayback Machine captures, with real @github counts back to 2009, and a daily script that logs a CSV.


Published · Updated · 7 min read

Cover image: GitHub's followers, 2009 to now

You can see a Twitter follower count history for free in the Wayback Machine, whose saved copies of public profiles show the count until mid-2022.

After that, you track the count yourself. This guide shows both, with real yearly counts for @github and a daily script for any public account. We run Weir, the API the script calls.

Paste this into your browser, swap in any username, and you get one saved copy of the profile per year:

https://web.archive.org/cdx/search/cdx?url=twitter.com/github&filter=statuscode:200&collapse=timestamp:4
  1. List the saved copiesThe CDX API gives one per year or month.
  2. Read each countOld profile pages print the exact number.
  3. Script the historyA short Python loop writes a CSV.
  4. Log today's count dailyOne profile lookup, 3 credits.
  5. Chart itOpen the CSV in Sheets or Excel.

Step 1: List the saved copies of a Twitter profile

The Wayback Machine is the Internet Archive's library of saved web pages. Its CDX API lists every saved copy of an address, with the date it was taken.

The link above sets three things:

  • url=twitter.com/github asks for the old address. In our test, x.com copies held no count.
  • filter=statuscode:200 keeps only copies that loaded.
  • collapse=timestamp:4 keeps one copy per year. Use timestamp:6 for one per month.

Each line is one capture, and the second column is its date and time:

com,twitter)/github 20090102090244 https://twitter.com/github text/html 200 ...
com,twitter)/github 20100406042636 https://twitter.com/github text/html 200 ...

To check one date by hand, open web.archive.org/web/*/twitter.com/github and pick a day on the calendar.

Step 2: See a past follower count on each capture

Open any capture from 2009 to mid-2022 and the count sits on the profile, just as it did that day. Later layouts show it short, like 1.7M, but its tooltip and the page's code hold the exact number.

We read every yearly capture of @github on 22 October 2026. The last bar is today's count from Weir:

@github followers, from Wayback captures and Weir
Jan 20091,073
Jan 201250,616
Jan 2015442,396
Jan 20181,704,786
Jun 20222,287,684
Oct 2026, via Weir2,724,229

The history tells a story the live profile can't. GitHub gained about 494,000 followers in 2017, then only about 74,000 the next year.

The trail stops in 2022, though. In our run:

  • Up to 2 June 2022, captures showed the exact count, 2,287,684 on that day.
  • A January 2023 capture showed only a rounded "2.4M".
  • From 2024 on, copies of both twitter.com/github and x.com/github held no count at all.

@NASA's captures broke at the same point: an exact count on 1 June 2022, none from July. So the archive covers the early years well and recent years poorly.

Step 3: Pull a Twitter follower count history with Python

A short script reads every yearly capture and writes a CSV. You need Python 3.8 or later and requests.

The first part finds the count in each old page layout, and lists the captures:

import re, sys, time, requests
PATTERNS = [r"follower_count[^>]*>\s*([\d,]+)",                  # 2009 to 2011
            r"([\d,]+)<span class='user-stats-stat'>Followers",  # 2012
            r"<strong>([\d,]+)</strong>\s*Followers",            # 2013
            r'title="([\d,.\s]+)[^"]*"\s+data-nav="followers"']  # 2014 to mid-2022

def count_in(html):
    hit = next((m for p in PATTERNS if (m := re.search(p, html))), None)
    return int(re.sub(r"\D", "", hit.group(1))) if hit else None

def captures(handle):  # one saved copy per year
    return requests.get("https://web.archive.org/cdx/search/cdx", timeout=90, params={
        "url": f"twitter.com/{handle}", "output": "json", "fl": "timestamp",
        "filter": "statuscode:200", "collapse": "timestamp:4"}).json()[1:]

The second part reads each capture. In our runs the archive dropped connections after about a dozen quick requests, so it waits 5 seconds between pages:

handle = sys.argv[1]
print("date,username,followers")
for (ts,) in captures(handle):
    day = f"{ts[:4]}-{ts[4:6]}-{ts[6:8]}"
    page = f"https://web.archive.org/web/{ts}id_/https://twitter.com/{handle}"
    try:
        count = count_in(requests.get(page, timeout=90).text)
    except requests.RequestException:
        print(day, "skipped: the archive is busy", file=sys.stderr)
        count = None
    if count:
        print(f"{day},{handle},{count}", flush=True)
    time.sleep(5)  # go easy on a free archive

Save both parts as wayback_followers.py and run python wayback_followers.py github > github_history.csv. Ours wrote, trimmed:

date,username,followers
2009-01-02,github,1073
2010-04-06,github,11901
2021-01-04,github,2016756
2022-01-02,github,2192730

Step 4: Track a Twitter follower count from today

From here on, you log the count yourself. Weir's GET /v1/x/users/{user} returns a public profile with its followersCount, for 3 credits: 1 for the call and 2 for the profile.

This script adds one row per account each time it runs. Set WEIR_API_KEY first:

import csv, datetime, os, sys, requests
HEADERS = {"Authorization": f"Bearer {os.environ['WEIR_API_KEY']}"}
def followers(username):
    r = requests.get(f"https://api.weirapp.io/v1/x/users/{username}",
                     headers=HEADERS, timeout=60)
    r.raise_for_status()  # 404: gone or suspended. 402: out of credits.
    return r.json()["data"]["followersCount"]

new_file = not os.path.exists("followers.csv")
with open("followers.csv", "a", newline="") as f:
    writer = csv.writer(f)
    if new_file:
        writer.writerow(["date", "username", "followers"])
    for username in sys.argv[1:]:
        writer.writerow([datetime.date.today(), username, followers(username)])

We looked up @github through Weir on 22 October 2026. It cost 3 credits and returned, trimmed:

{
  "username": "github",
  "followersCount": 2724229,
  "followingCount": 334,
  "createdAt": "2008-02-11T04:41:50.000Z"
}

Run the script once a day. On Linux or a Mac, a cron line such as 0 9 * * * runs it every morning at 9; on Windows, use Task Scheduler.

Here's what a daily check costs:

3credits per account per day
93credits a month at most, per account
53accounts a day on 5,000 free credits

Tracking many accounts? GET /v1/x/users takes up to 100 numeric ids in one call, at 1 credit plus 2 per profile.

Step 5: Chart your Twitter follower growth

Both CSVs share the same three columns. Paste them into one sheet in Google Sheets or Excel, sort by date, and insert a line chart.

Expect a gap between mid-2022 and the day you started logging. The archive won't fill it for most accounts, so start the daily log today.

What does X's own analytics show about follower history?

Less than you'd hope, going by Wayback copies of X's help pages:

  • X's analytics help covers post and video stats, 28 days by default, and says nothing about follower history.
  • X's Premium page lists its features by tier and doesn't mention analytics.
  • Your X data download includes "a list of your followers," a snapshot rather than a count over time.

Postory's October 2026 guide says X's Premium dashboard shows "daily numbers of follows and unfollows." We couldn't confirm that on X's pages, and it would cover only your own account.

X's API has no history either: its data dictionary lists one followers_count per user. For what X does show for free, see Twitter analytics without Premium.

If it doesn't work

  • No captures come back. Nobody saved that profile, or it used another handle. Our username history guide helps find old handles.
  • A capture has no count. It's from mid-2022 or later, or the saved page is an empty shell. Try a nearby date.
  • The script skips a year. If it printed "skipped", the archive was busy. Wait a few minutes and run it again.
  • Weir returns a 404. The account is suspended or no longer exists.

When not to use Weir

  • You only need past counts. The Wayback Machine is free and needs no key.
  • You check once a month. Look at the profile and write the number down.
  • You need who followed, not how many. A count can't show that; see how to export Twitter followers to CSV.
  • Cost per lookup matters most. TwitterAPI.io charges less per profile, and far less in batches of 100 or more.

Track brands and organisations, not private people. X's Terms of Service say you "cannot scrape the Services without X's express written permission," so check what applies to you. This isn't legal advice.

FAQ

Can you see past follower counts on Twitter?

Not on X itself, because a profile shows only today's count. Saved copies in the Wayback Machine show the exact count for most years up to mid-2022.

How do I track my Twitter follower count over time?

Log it once a day. A script that calls Weir's profile lookup and appends the count to a CSV costs 3 credits a day. Weir's 5,000 free monthly credits cover about 50 accounts.

Is there a free Twitter follower growth tracker?

For past years, the Wayback Machine and the script above cost nothing. For daily tracking, Weir's free monthly credits cover about 50 accounts. Some paid tools estimate the time before you signed up: Fedica says it "estimates what your history probably looked like."

Does X analytics show follower history?

X's own help pages don't say so. They describe post stats over 28 days by default. Postory's guide says Premium's dashboard shows daily follows and unfollows for your own account, which we couldn't confirm on X's pages.

Sources

All checked on 22 October 2026. X's pages blocked direct reads, so we read Wayback copies, dated in brackets:

Keep reading