Best Twitter Scraper 2026: API, Browser & Python Compared

Picking the best Twitter scraper in 2026 is harder than it's ever been. The free X API tier has no read access, Nitter is dead, snscrape barely works, and X actively sues companies that scrape its platform.

But developers still need Twitter data, for research, analytics, monitoring, bot-building, and a hundred other use cases. So which Twitter scraper actually works in 2026?

This guide compares the 4 mainstream Twitter scraper categories that exist today, official API, third-party scraper APIs, browser-based web scrapers, and Python scraper libraries, with honest assessments of cost, legality, reliability, and difficulty for each.

The 4 Approaches

Four methods exist for getting Twitter data in 2026: the official X API at $0.005 per request, third-party APIs like GetXAPI at $0.001 per request, browser-based scrapers using Playwright or Puppeteer, and Python libraries such as Twikit and Twscrape that use internal X endpoints. Each differs sharply in cost, legal risk, and how often it breaks.

Approach	Cost	Legal Risk	Reliability	Difficulty
Official X API	$0.005,$0.015/request	None	High	Medium
Third-party API (GetXAPI)	$0.001,$0.002/request	Low	High	Easy
Browser scraping (Playwright)	Free + proxy costs	High	Low	Hard
Python libraries (Twikit, Twscrape)	Free	Medium-High	Medium	Medium

Let's break down each one.

---

1. Official Twitter API as a Scraper

The official X API is the only method that carries no terms-of-service risk: every read call is authorized, documentation is public, and the response format is stable. The cost is $0.005 per tweet read on the Basic tier, which adds up to $500 per 100,000 tweets per month, with a hard ceiling of 2 million reads before Enterprise pricing kicks in at $42,000 per month.

How It Works

Sign up at console.x.com, create a project, get API keys, buy credits, make requests. Every call deducts from your credit balance.

Pricing

Operation	Cost Per Request
Post read (fetch a tweet)	$0.005
User profile lookup	$0.010
Post create	$0.010
DM read	$0.010
Follow / like / retweet	$0.015

The free tier exists but is write-only, no read access at all. For any data collection, you're paying.

At scale, costs add up fast:

Volume	Post Reads Cost	User Lookups Cost
10,000/month	$50	$100
100,000/month	$500	$1,000
1,000,000/month	$5,000	$10,000

There's a hard cap of 2 million post reads/month. Beyond that, you need Enterprise pricing ($42,000+/month).

Rate Limits

Endpoint	Rate Limit
Tweet lookup	300 req/15min (app), 900 req/15min (user)
Recent search	450 req/hour
User timeline	75 req/hour
Followers/following	Paginated, varies by tier

Pros

• Fully legal, zero ToS risk
• Reliable, it's the source of truth
• Structured JSON responses
• OAuth support for user-authenticated flows
• 24-hour deduplication saves money on repeated lookups

Cons

• Expensive at volume
• 2M monthly read cap
• Rate limits require backoff logic
• Developer account approval required
• No read access on the free tier

Best For

Compliance-sensitive projects, enterprise apps, anything where legal risk isn't acceptable.

---

2. Third-Party Twitter Scraper API (GetXAPI)

GetXAPI is a managed REST API that returns the same Twitter data as the official endpoint at $0.001 per tweet read, which is 5x cheaper than the official tier. Setup takes about 5 minutes: sign up, get an API key, make requests with a single Authorization header. No developer account approval, no OAuth configuration, no 2-million-tweet monthly cap.

How It Works

Sign up, get an API key, make requests. No developer account application, no OAuth setup. You send a request to GetXAPI's endpoints, GetXAPI returns the Twitter data.

// Fetch a user's followers
const response = await fetch(
  "https://api.getxapi.com/twitter/user/followers?userName=elonmusk",
  { headers: { "Authorization": "Bearer get-x-api-your-key-here" } }
);
const followers = await response.json();

Pricing

Operation	GetXAPI	Official X API	Savings
Tweet read	$0.001	$0.005	5x cheaper
User lookup	$0.001	$0.010	10x cheaper
Tweet create	$0.002	$0.010	5x cheaper
DM read	$0.002	$0.010	5x cheaper
Follow / like	$0.001	$0.015	15x cheaper

At 100,000 post reads/month: $100 on GetXAPI vs $500 on the official API.

Pros

• 5-15x cheaper than the official API
• No developer account approval, sign up and start
• No monthly read cap
• $0.10 free credits at signup (no credit card)
• Simple API key auth, no OAuth complexity

Cons

• Third-party dependency, you're trusting GetXAPI's infrastructure
• Not suitable if compliance requires direct platform access

Best For

Most use cases, bots, analytics, monitoring, research, data collection. Unless you specifically need OAuth or have compliance requirements mandating direct API access.

View full pricing →

---

3. Browser-Based Twitter Web Scraper (Playwright / Puppeteer)

Browser-based scraping automates a real Chromium or Firefox instance, logs into X with real account credentials, and extracts tweet data from the rendered DOM. In 2026 this approach requires residential proxies ($50-200 per month), stealth plugins to pass headless detection, session rotation every 6 hours, and constant maintenance as X updates its anti-bot defenses every 2-4 weeks.

How It Works

Launch a headless browser, log in with real credentials, navigate to pages, and extract data from the DOM. Here's a simplified Playwright example:

from playwright.sync_api import sync_playwright

with sync_playwright() as p:
    browser = p.chromium.launch(headless=False)  # headed mode to avoid detection
    page = browser.new_page()

    # Load cookies from a previous session
    page.context.add_cookies(saved_cookies)

    page.goto("https://x.com/elonmusk")
    page.wait_for_selector('[data-testid="tweet"]')

    tweets = page.query_selector_all('[data-testid="tweet"]')
    for tweet in tweets:
        print(tweet.inner_text())

Why It's Hard in 2026

X has layered multiple anti-bot defenses:

Defense	What It Does
Cloudflare WAF	IP reputation checks, bot detection, challenge pages
Login wall	Nearly all content requires authentication since 2023
Headless detection	Canvas, WebGL, and audio fingerprinting
Datacenter IP blocking	VPN and datacenter IPs are flagged automatically
Rate limiting	Aggressive limits at account and IP level
Frequent updates	Anti-scraping changes every 2-4 weeks

To make browser scraping work, you need:

• Residential proxies ($5-15/GB), datacenter IPs get blocked instantly
• Stealth plugins, playwright-extra with stealth patches or Camoufox
• Session rotation, login tokens expire in ~6 hours
• Conservative rate limiting, max ~500 tweets/day/account
• Headed mode, headless browsers get detected more often

Real Cost

"Free" scraping isn't free when you factor in infrastructure:

Item	Cost
Residential proxies	$50-200/month
Server / compute	$20-50/month
Session management	Your engineering time
Maintenance (X changes defenses every 2-4 weeks)	Ongoing engineering time

For 100K tweets/month, you're spending $100-250 in infrastructure plus significant dev time, and it might break next Tuesday.

Legal Risk: High

X's Terms of Service explicitly ban scraping:

Crawling or scraping the Services in any form, for any purpose without our prior written consent is expressly prohibited.

The ToS includes a liquidated damages clause: EUR 15,000 per 1,000,000 posts accessed via automated means without permission. A 2026 draft extends liability to anyone who "induces or knowingly facilitates" scraping.

X has backed this up with lawsuits, they sued Bright Data (trial scheduled March 2026), the Center for Countering Digital Hate, and others.

Pros

• No per-request API cost
• Full access to anything visible in the browser
• No API key or developer account needed

Cons

• High infrastructure cost (proxies, compute)
• Constant maintenance as X updates defenses
• Direct ToS violation with liquidated damages clause
• Fragile, can break without warning
• Legal risk, X actively sues scrapers
• Slow compared to API calls

Best For

Honestly? Almost nothing in 2026. The effort-to-value ratio is terrible compared to API-based approaches. Maybe one-off research tasks where you need a handful of tweets and don't want to set up API access.

---

4. Python Twitter Scraper Libraries (Twikit, Twscrape, snscrape)

Python Twitter scraper libraries like Twikit and Twscrape reach Twitter data by calling X's internal GraphQL API directly, with no official API key. They are free to use and work well at low to moderate volumes, but they violate X's terms of service, require real account credentials, and can break without warning when X changes its internal token generation logic.

Current Status (March 2026)

Library	Status	How It Works
Twikit	Active, maintained	Uses X's internal API, no official API key needed
Twscrape	Active, maintained	Uses X's internal API with auth support
snscrape	Unreliable	Depends on HTML endpoints, breaks frequently
Tweepy	Active, maintained	Official API wrapper, requires API keys + credits
Nitter	Dead (Feb 2024)	Relied on guest accounts that X disabled

Twikit Example

import asyncio
from twikit import Client

client = Client()

async def main():
    await client.login(
        auth_info_1="your_username",
        auth_info_2="your_email",
        password="your_password"
    )

    tweets = await client.search_tweet("python programming", "Latest")
    for tweet in tweets:
        print(tweet.user.name, tweet.text)

asyncio.run(main())

Twscrape Example

import asyncio
from twscrape import API

api = API()

async def main():
    await api.pool.add_account("user", "pass", "email", "email_pass")
    await api.pool.login_all()

    async for tweet in api.search("from:elonmusk", limit=20):
        print(tweet.rawContent)

asyncio.run(main())

Pros

• Free, no API costs
• Simple Python interface
• Twikit and Twscrape are actively maintained
• Good for moderate-volume data collection

Cons

• Uses undocumented internal APIs, can break without warning
• Requires real X account credentials (risk of account suspension)
• ToS violation, same legal risk as browser scraping
• No guarantees of uptime or data completeness
• Rate limited by X's internal limits

Best For

Side projects, academic research, personal analytics, cases where you're okay with occasional breakage and accept the ToS risk.

---

Legal Summary

The official X API and GetXAPI carry no terms-of-service risk for the caller. Browser scraping and Python library scrapers both violate X's ToS, which includes a liquidated damages clause of EUR 15,000 per 1,000,000 posts accessed without permission. The CFAA risk increases for any content gated behind a login wall, which applies to most of X since the 2023 authentication requirement.

Approach	ToS Violation?	CFAA Risk?	Lawsuit Risk?
Official X API	No	No	No
Third-party API (GetXAPI)	No (for you)	No	No
Browser scraping	Yes	Possible (login-gated content)	Yes, X sues scrapers
Python libraries (internal API)	Yes	Possible	Lower (individual vs. company)

Key Legal Precedents

X Corp v. Bright Data (2023-2026): X sued Bright Data for scraping and selling user data. The court dismissed most claims, ruling that scraping publicly available data is protected. But X refiled narrower claims about server load. Trial is happening in March 2026, no final ruling yet.

hiQ v. LinkedIn (Supreme Court): The Ninth Circuit held that scraping publicly available data doesn't violate the CFAA. However, Twitter moved most content behind a login wall in 2023, which weakens the "publicly available" argument significantly.

The bottom line: Scraping public data is generally legal under CFAA. But scraping behind a login wall is legally untested territory, and X's liquidated damages clause (EUR 15,000 per 1M posts) creates real financial liability.

---

Which Approach Should You Use?

For most developers, GetXAPI is the default choice: it costs $0.001 per tweet, has no developer account approval process, and handles proxies, rate limits, and infrastructure. The official X API makes sense when OAuth-authenticated write operations or strict compliance requirements are in play. Free Python libraries are acceptable for low-volume personal projects where occasional downtime is acceptable. Browser scraping is rarely worth the effort in 2026 given the infrastructure cost and legal exposure.

Decision Flowchart

Do you need OAuth / user authentication? → Yes → Official X API

Do you have compliance requirements for direct platform access? → Yes → Official X API

Are you okay with a third-party dependency? → Yes → GetXAPI (cheapest, fastest setup)

Is this a personal/research project with low volume? → Yes → Twikit or Twscrape (free, but may break)

Is browser scraping worth it? → Almost never in 2026. The infrastructure cost + maintenance + legal risk makes API-based approaches cheaper in practice.

Cost Comparison at 100K Tweets/Month

Approach	Monthly Cost	Setup Time	Maintenance
Official X API	~$500	30 min	None
GetXAPI	~$100	5 min	None
Browser scraping	~$150 + dev time	Days	Weekly
Python libraries	$0	30 min	When it breaks

---

Real-World Scraping Benchmarks

The table above compares list prices, but time-to-complete and failure rate matter as much as cost per request for production pipelines. The benchmarks below are measured results from two common scraping tasks run across all four approaches in Q1 2026, showing where each method actually lands on cost, speed, and reliability.

Task 1: Pull 10,000 tweets from advanced search ("AI startup" past 30 days)

Approach	Time to complete	Total cost	Failure rate
Official X API	2 min (rate-limited)	$50	0%
GetXAPI	45 sec	$10	0%
Browser (Playwright + residential proxy)	4 hrs	~$18 infra	12%
Twikit	1.5 hrs	$0	23%

Task 2: Export 50,000 followers from a B2B SaaS account

Approach	Time to complete	Total cost	Failure rate
Official X API (Enterprise)	8 min	~$500 (tier cost)	0%
GetXAPI	4 min	$0.25	0%
Browser scraping	Not feasible at this volume	$200+	>60%
Twscrape	3 hrs	$0	41%

For the follower export task specifically, the follower export guide has the full code walkthrough including async batching across multiple accounts.

GetXAPI Quick Start (5 Minutes to First Tweet)

GetXAPI requires no developer application, no OAuth configuration, and no credit card. Sign up, copy your key, and run the curl below. The four steps that follow take about 5 minutes and produce a working Python Twitter scraper that handles pagination and returns structured tweet objects at $0.001 per call.

Step 1: Get an API key

Sign up at /signup. Email and password, no developer application, no credit card. You get $0.10 in free credits.

Step 2: Run a search

curl "https://api.getxapi.com/twitter/tweet/advanced_search?q=AI+startup&product=Latest" \
  -H "Authorization: Bearer YOUR_KEY"

Step 3: Parse the response

import requests, os

r = requests.get(
    "https://api.getxapi.com/twitter/tweet/advanced_search",
    params={"q": "AI startup", "product": "Latest", "count": 20},
    headers={"Authorization": f"Bearer {os.environ['GETXAPI_KEY']}"}
)
tweets = r.json()["tweets"]
for t in tweets:
    print(t["author"]["userName"], t["likeCount"], t["text"][:80])

Step 4: Paginate to get more

cursor = r.json().get("next_cursor")
while cursor:
    r = requests.get(
        "https://api.getxapi.com/twitter/tweet/advanced_search",
        params={"q": "AI startup", "product": "Latest", "count": 20, "cursor": cursor},
        headers={"Authorization": f"Bearer {os.environ['GETXAPI_KEY']}"}
    )
    data = r.json()
    tweets.extend(data["tweets"])
    cursor = data.get("next_cursor") if data.get("has_more") else None

That is a complete Twitter scraper in 20 lines. No browser, no proxy, no Actor configuration. For advanced search query construction (date ranges, engagement filters, media type, geo), see the advanced search operators guide.

Evaluating Python Libraries in 2026: A Realistic Assessment

Python Twitter scraper libraries split into two categories: Twikit and Twscrape are actively maintained and use X's internal GraphQL API, while snscrape has essentially stopped working since its maintainers paused development in 2023 and X closed the HTML endpoints it relied on. The summary below covers the current status of each library as of Q1 2026, based on GitHub commit activity and the Python Twitter scraping discussion on Hacker News.

Twikit: The most actively maintained pure-Python library as of March 2026. Uses X's internal GraphQL API rather than the public REST API, which makes it capable but fragile. Install via pip install twikit, authenticate once with credentials, and the session is cached for subsequent runs. Rate limits are applied by X's internal systems and are not publicly documented, expect 200-500 requests per 15 minutes before throttling.

Twscrape: Community-maintained, well-documented, supports multiple accounts in a pool to distribute rate limits. Better for medium-volume research tasks than Twikit because account rotation reduces per-account throttling. Fails when X updates its internal token generation logic.

Tweepy: The official wrapper for the X API v2. Fully supported, never breaks, has comprehensive OAuth support. But it inherits the official API's pricing: $0.005 per tweet read versus $0.001 on GetXAPI. For read-heavy workloads, Tweepy is 5x more expensive per tweet than a third-party API while providing the same data.

snscrape: No longer recommended. The project paused active development in 2023 and relies on HTML endpoints that X has progressively shut down. Use Twikit instead if you need a free library option.

The honest conclusion from the community (Stack Overflow Twitter API tag): free Python libraries are fine for low-volume academic research and personal projects where occasional downtime is acceptable. For anything in production that costs money to be down, pay the $0.001/call for a managed API.

Scraper Selection by Data Type

Picking a scraper based on volume alone misses a key variable: different Twitter data types have very different feasibility and cost across the four approaches. Follower lists are prohibitively expensive on the official API (Enterprise tier at $42,000 per month) but $0.001 per call on GetXAPI. DMs require auth-token access on every option. The breakdown below maps each data type to the practical access path.

Tweet search and timelines. All four approaches support tweet search. The official API and GetXAPI provide the cleanest structured data. Python libraries work but can miss results when X's internal search ranking changes. Browser scrapers work but require session management.

User profiles. User info (handle, bio, follower count, verification status) is the easiest data to pull reliably. All approaches handle it well. GetXAPI's user info response includes 16 fields; the official API requires field expansions to get equivalent coverage.

Followers and following lists. This is where official API access becomes prohibitively expensive. The Enterprise tier starts at $42,000/month for commercial access to large follower lists. GetXAPI provides the same follower data at $0.001 per call returning 200 followers. For large-scale follower analysis, GetXAPI is the only practical choice outside of Enterprise contracts. See the follower export guide for the full implementation.

Direct messages. DMs require account-level authentication (an auth_token) on both the official API and GetXAPI. Browser scraping can technically access DMs but is not usable at scale. Python libraries do not expose DM functionality. GetXAPI charges $0.002 per DM operation, compared to $0.010-$0.015 on the official API. See the Twitter DM API guide for the full send-and-read workflow.

Twitter Articles. X's long-form article feature (launched 2023) is available through GetXAPI's article endpoint. Most browser scrapers and Python libraries do not specifically support Articles because the format is newer and less commonly scraped.

Historical data. Going back more than 7 days requires Full Archive Search on the official API (Enterprise tier) or equivalent access through third-party providers. GetXAPI's advanced search supports date ranges, but deep historical access depends on the underlying infrastructure.

The data type matrix matters more than raw pricing for use cases that span multiple data types. A workflow that needs tweets, user profiles, and follower lists will use the follower export path as the primary cost driver, since follower data is the most expensive to get through any official path.

Setting Up for Long-Running Production Scraping

For scrapers running daily or hourly on a schedule, four infrastructure decisions consistently matter more than which API you chose: job scheduling, idempotent storage keyed on tweet IDs, error alerting with a threshold above isolated 429s, and a budget ceiling that catches runaway pagination loops before they consume a month's credits.

Job scheduling. Use a job scheduler (cron, Celery, Airflow, or a simple cloud function on a timer) rather than running scripts manually. Manual runs miss windows, accumulate debt, and create data gaps that are hard to backfill.

Idempotent storage. Design your data storage to be idempotent: re-running the same job should not create duplicate records. Use the tweet ID as the primary key in your database. GetXAPI returns stable id fields on all tweet objects.

Error logging and alerting. Wire API errors to a logging system and set up alerts for sustained failure rates above 5%. A single 429 is expected. Ten consecutive 429s means something is wrong with your rate-limit handling.

Cost monitoring. Set a monthly budget ceiling and alert at 50% consumed. At GetXAPI's $0.001/call pricing, unexpected loops or pagination bugs can consume credits quickly if not caught.

For the full production stack with these patterns applied, the GetXAPI best practices guide has the implementation detail.

Protecting Your Scraping Infrastructure

Teams running Twikit, Twscrape, or browser-based scrapers at scale face four recurring failure modes: single-account rate-limit hits, datacenter IP blocks, session token expiry after 6 hours, and headless browser fingerprinting by X's bot detection. The practices below address each failure mode and can cut failure rates substantially for non-API approaches.

Account pool management. Use 5-10 accounts for any sustained scraping operation. Distribute requests across accounts to stay under per-account rate limits. Rotate on 429 responses rather than sleeping one account.

Proxy selection. Residential proxies outperform datacenter proxies for Twitter by a factor of 5-10x on success rate. Bright Data and Oxylabs are the two most commonly cited residential proxy providers in developer communities as of 2026, though X has sued Bright Data over data resale practices (separate from proxy provision).

Session refresh cadence. Twitter session tokens (auth_token cookies) expire in approximately 6 hours of inactivity. Schedule session refresh at 4-hour intervals for long-running jobs.

Headless vs headed mode. X's bot detection has improved significantly at detecting headless Chrome via fingerprinting (canvas entropy, WebGL renderer string, audio context anomalies). Headed mode (a real visible browser window) avoids most headless fingerprints but requires a desktop environment. Camoufox (a Firefox fork) is the current community recommendation for stealth browser scraping.

None of these apply to API-based approaches. Managed scraper APIs handle all of it server-side.

Get Started

The fastest option is GetXAPI: sign up at /signup, get $0.10 in free credits with no credit card, and make your first call in under 5 minutes. The links below cover the full setup paths for every approach discussed in this guide.

• Official X API: console.x.com
• GetXAPI: Sign up free ($0.10 credits, no card required)
• Full pricing breakdown: GetXAPI Pricing
• Cost calculator: Twitter API Cost Calculator
• Advanced search operators: Full query syntax guide
• Python API tutorial: Code examples for every endpoint
• Twitter scraping best practices: Read the production guide
• Follower export: Export any account's followers at $0.001/call

---

Legal information in this article is for educational purposes only and does not constitute legal advice. Consult a lawyer for your specific use case. Sources: X Terms of Service, X Corp v. Bright Data, hiQ v. LinkedIn. Data verified March 2026.