Guide
How to scrape Twitter without the API
The X API used to be the obvious way to get Twitter data. Since 2023 it's priced for enterprises, and everyone else — researchers, marketers, indie hackers — has been looking for the exit. These are the four exits that actually exist, with the trade-offs of each.
Updated August 2026 · Public data only — followers, following, tweets, search
Why everyone avoids the API
The free tier is essentially write-only — it exists so bots can post. The Basic tier costs $200/month with caps too tight for real collection work, and the tiers with genuine read access run from thousands per month. For a one-off follower list or a monthly brand search, the official route prices itself out of the conversation.
Meanwhile the data itself is public — anyone can open a profile and read it. "Scraping without the API" just means collecting that public data some other way. The four ways differ in who does the work and whose account carries the risk.
Option 1: Browser extensions
Chrome extensions scroll the page for you and scrape what renders. Zero setup, and they look free. The catch is where they run: inside your logged-in session. X rate-limits the scrolling almost immediately, so large lists take hours in a tab you can't touch and often stall partway. Worse, the automation is indistinguishable from you misbehaving — it's your account that gets flagged, locked, or suspended.
Fine for a few hundred rows on an account you could afford to lose. Wrong tool for anything that matters — the full argument is in the follower-export guide.
Option 2: DIY scripts
The engineering route: a headless browser (Playwright, Puppeteer) or reverse-engineered internal endpoints, plus proxies, plus retry logic. It's the most flexible option and the most expensive in the only currency that matters — maintenance. X changes its frontend and anti-bot systems constantly; every change breaks your scraper on a schedule you don't control. The old shortcuts are gone too: guest access ended in 2023, which killed Nitter and the popular unauthenticated Python libraries. Serious DIY scraping now means managed sessions and proxy pools — a part-time infrastructure job.
Worth it if data collection is your product. Not worth it if you just need the data.
Option 3: Scraping APIs
Commercial scraping APIs (Apify actors and similar) run the infrastructure and sell you the output by the thousand rows. Reliability is good, your account is never involved, and the pricing lands far below X's official tiers. The catch: they are developer products. You're still writing code, handling pagination and JSON, and wiring results into a file — the API problem at a fairer price, not the absence of the problem.
Option 4: No-code exports
An export service is the same idea taken one step further: the infrastructure and the pipeline, behind a form. TweetExporter takes a handle or a search query, shows the exact price up front, and delivers CSV, XLSX, or JSON — one row per follower, followed account, or tweet, with full public metadata. You never log in with X, so there's no session to endanger, and the first 50 rows are free.
The trade-off is scope: you get the four export types — followers, following, tweets, and search results — not arbitrary custom crawls. For most people that's the whole shopping list.
The four options, compared
| Method | Effort | Cost | Reliability | Account risk |
|---|---|---|---|---|
| Browser extensions | Low | Free–$20/mo | Poor — rate-limited, stalls on big lists | High — runs as your account |
| DIY scripts | High — code + constant upkeep | Your time + proxies | Fragile — breaks with every X change | High if logged in |
| Scraping APIs | Medium — still code | $50–500+/mo | Good | None |
| No-code export (TweetExporter) | None — a form | First 50 rows free, then credits | Good — managed infrastructure | None — no X login |
What data you can get
Everything public, nothing private. A follower or following export delivers full profile rows — handle, name, bio, follower counts, location, website, join date. A tweet export delivers full text, timestamps, engagement counts, and permanent links, with threads reassemblable by reply ID. A search export turns any query X search accepts — including the full operator language — into a dataset. DMs, likes, and protected accounts are out of reach for every method on this page; that data isn't public.
Scraping without the API, answered
Is scraping Twitter data legal?+
Collecting public data — what anyone can see in a browser — is broadly treated differently from breaching private data, but platform terms and local laws still apply, and you're responsible for how you use what you collect. TweetExporter only reads public data and respects X's rate limits.
Will my X account get banned for scraping?+
It can, if the scraping runs as your account — that's the core risk of browser extensions and logged-in scripts. Automation that X's systems attribute to your session puts your session on the line. An export service that never touches your login carries no such risk to you.
Why not just use the free API tier?+
Since 2023 the free tier is essentially write-only — it exists for posting bots. Reading followers, timelines, or search at any useful volume starts at paid tiers, and the tiers with real read access cost hundreds to thousands per month.
What Twitter data can be exported without the API?+
Anything public: an account's followers, who they follow, their full tweet history, and the results of any search query. Each comes out as one row per item with full public metadata, in CSV, XLSX, or JSON.
What happened to Nitter and the old scraping tricks?+
Most died in 2023–2024 when X removed guest access. Nitter instances went dark, and the Python libraries that depended on unauthenticated endpoints broke. Everything that still works today either pays for the API, runs a logged-in session, or operates managed infrastructure the way an export service does.
Do I need to know how to code?+
Only for the DIY route. A no-code export is a form: type a handle or a search query, see the price, download the file. If your goal is the data rather than the pipeline, you never touch code.