A Twitter scraping API is a hosted service that returns X (Twitter) data as structured JSON over HTTP, so you never run browsers, proxies or account pools yourself. Six options compete in 2026: TwitterAPIs, ScrapeBadger, Apify Tweet Scraper, Scrapingdog, Bright Data, and self-hosted open-source libraries. TwitterAPIs sits at the bottom of the no-commitment price chart at $0.04 per 1,000 tweets, billed $0.0008 per call with no subscription and no platform rate ceiling. Credit subscriptions can price a request lower once you commit to a large monthly plan, so the split is simple: metered billing with no floor favours TwitterAPIs, and heavy committed volume can favour ScrapeBadger or Scrapingdog.
How we priced this ranking
Written by Emma, twitterapis developer relations
Every third-party rate below was read from that vendor's own live pricing or product page on 12 August 2026, not from another comparison article. Our own read rate is $0.0008 per call (source: our published pricing), which is $0.04 per 1,000 tweets. Where a vendor publishes a plan price and a credit allowance rather than a rate, we divide one by the other and label the result as effective rather than quoted.
Two figures on this page are derived, not published: Bright Data's per-record rate, and every effective per-1,000 rate computed from a plan price. Both are marked where they appear.
What the meter counts
A credit, a request, a result and a tweet are four different units. Normalize every vendor to the same one before comparing, or the cheapest-looking plan wins on arithmetic alone.
What the floor costs
A subscription bills the same whether you send one request or a million. Metered billing does not. Which is cheaper depends entirely on how steady your volume is.
What you can test for free
A free tier that blocks API access, or covers 40 requests, cannot validate an integration. Check what the free plan actually permits before you budget for the paid one.
ScrapeBadger spends 1 credit per Twitter request and prices credits from $0.150 per 1,000 pay as you go down to $0.054 per 1,000 on its $699 Scale plan. (ScrapeBadger pricing, 2026)
Scrapingdog charges 5 credits per Twitter Scraper API request, so its $40 LITE plan covers about 40,000 Twitter requests, an effective $1 per 1,000. (Scrapingdog pricing, 2026)
Bright Data includes 5,000 free records a month and values them at about $7.50, which works out to roughly $1.50 per 1,000 records. (Bright Data Twitter scraper, 2026)
The Lineup at a Glance
Provider
Billing Model
Entry Point
Free Tier
Best For
Tradeoff
TwitterAPIs
Metered per call, no subscription
$0.0008 per call (about 20 tweets)
$0.50 in credits, no card
Teams that want spend to track usage with no monthly floor
A committed credit subscription can beat it on price per request at very high volume
ScrapeBadger
Credits, pay as you go or monthly
$0.15 per 1,000 credits, 1 credit per request
1,000 credits on signup, no card
High committed monthly volume where the per-request rate is what matters
Page size per credit is not published, so cost per tweet cannot be compared
Apify (Tweet Scraper V2)
Pay per result, on a platform plan
$0.40 per 1,000 tweets
Demo only, 5 runs a month, no API access
Extraction that already lives inside an Apify workflow
The free plan cannot call the actor over the API at all
Scrapingdog
Monthly subscription, credit metered
$40 a month, 5 credits per Twitter request
200 credits, about 40 Twitter requests
Buyers who want one vendor across Twitter, Google, LinkedIn and more
The 5x credit multiplier means a plan buys a fifth of the requests it advertises
Bright Data
Per delivered record, plan or pay as you go
About $1.50 per 1,000 records, derived from their own free-tier valuation
5,000 records a month, no card
Enterprise buyers who need procurement, SLAs and unlimited concurrency
The published per-record rate is rendered as an animated widget, not a readable number
Self-hosted open source
Free software, you pay for the infrastructure
$0 in licence, proxies and accounts on you
Entirely free to install
Research, one-off pulls, and anyone who enjoys maintaining a scraper
Breaks whenever X changes, and you absorb every ban and CAPTCHA yourself
Heads up: scraping vendors reprice often, and several of them bill in credits rather than in tweets. Before you commit budget, reconfirm the current numbers on each provider's own pricing page and check what unit their meter counts. Every figure here was read from source on 12 August 2026.
Where we win, and where we do not
A ranked list written by one of the vendors on it is worth reading only if it says where the vendor loses. Here is ours. TwitterAPIs is the cheapest pay-as-you-go option on this page per tweet, metered with no monthly commitment: $0.0008 a call returning about 20 tweets is $0.04 per 1,000, against $0.40 on the headline Apify actor and about $1 per 1,000 requests on Scrapingdog's entry plan.
Now the other direction. ScrapeBadger sells credits, spends exactly one on a Twitter request, and prices its $699 Scale plan at an effective $0.054 per 1,000 credits. Measured per request, that is well under our $0.0008 per call. If your volume is genuinely large and genuinely steady, that plan is cheaper per request than metering with us, and we would rather write that down than leave it out.
The comparison stops there, though, and the reason matters. A credit buys a request, and neither ScrapeBadger nor Scrapingdog publishes how many tweets a Twitter request returns. Our own call returns about 20. Without their page size, a cost-per-tweet comparison cannot be computed from published figures, only a cost-per-request one. We have not estimated it, because an estimate presented next to measured numbers reads as measured.
So the practical rule: if your volume swings, is seasonal, or is still being sized, metered billing with no floor wins on total spend. If you can commit to millions of requests a month for the foreseeable future, price a credit plan against us on your own query, count the rows each returns, and buy on that number rather than on either vendor's headline.
6 Twitter Scraping Options, Card by Card
Each card lays out the billing model, what the meter counts, the endpoint surface, the auth shape and the free-tier terms, then names the workloads it fits and the ones it does not.
The cheapest metered rate per tweet in this set, billed purely per call, with no subscription to clear before the first request.
Billing Model
Strictly per call. A read endpoint costs $0.0008 and returns about 20 tweets, so 1,000 tweets works out to $0.04. There is no monthly plan, no spend floor, and no platform-imposed rate ceiling.
$0.0008 per call, about 20 tweets per standard response
$0.04 per 1,000 tweets on the read tier
The same $0.0008 rate holds across tweets, users, search, lists and the follower graph
New accounts open with $0.50 in credits, roughly 625 calls, and no card on file
Endpoint Surface
99 endpoints, 63 reads plus 36 writes, covering search, timelines, threads, profiles, lists and followers
Auth Model
One bearer-token header, minted in the dashboard
Free Tier
$0.50 in starter credits, no card, no expiry clock
Best Fit
Workloads whose monthly volume swings, where a fixed plan turns into dead spend
Solo builders and side projects that cannot justify a $40 to $700 monthly floor
Pipelines that need the follower graph, thread expansion or write actions, not just tweet pulls
Anyone who wants one invoice line that maps cleanly onto API calls
Worst Fit
Buyers contractually required to source data directly from X Corp
Very large committed monthly volume, where a credit subscription can beat the per-request rate
Verdict
The default pick for most teams searching this term. It leads on price per tweet with no commitment, carries the broadest endpoint surface here, and needs one header to authenticate.
Read-heavy work belongs here by default: search, timelines, user lookups and the follower graph all bill at the same flat rate. The place it loses is named honestly in the section below rather than left out.
A credit-based scraping platform whose Twitter endpoint costs a single credit per request, and whose largest plan prices credits lower than anyone else here.
Billing Model
Credits, bought either pay as you go or on a monthly plan. The Twitter Scraping API spends 1 credit per request, so a plan's credit count is also its request count. Pay-as-you-go credits never expire; subscription credits refresh monthly.
Pay as you go from $10, at $0.150 per 1,000 credits
Starter $49 a month for 600,000 credits, an effective $0.082 per 1,000
Growth $129 a month for 1,800,000 credits, an effective $0.072 per 1,000
Scale $699 a month for 13,000,000 credits, an effective $0.054 per 1,000
The Twitter Scraping API spends 1 credit per request
Endpoint Surface
Twitter search and profile scraping plus a separate Twitter Streams product for WebSocket and webhook monitoring, alongside Google, Reddit, Amazon, Instagram, YouTube and TikTok endpoints
Auth Model
An API key in a single header
Free Tier
1,000 credits on signup, no credit card required
Best Fit
Heavy, predictable monthly volume that can commit to a large plan
Teams that also scrape Reddit, Amazon or Google and want one credit pool
Real-time account and keyword monitoring, which their Streams product covers directly
Worst Fit
Anyone comparing cost per tweet, because the tweets returned per credit are not published
Bursty or seasonal work, where a monthly credit refresh expires unused capacity
Verdict
The strongest challenger here on raw price per request. At the Scale tier a credit costs a fraction of a metered call, and we would rather point that out than pretend the chart only goes one way.
The catch is a measurement one, not a marketing one. A credit buys a request, and their pages do not state how many tweets a Twitter request returns, so a like-for-like cost per tweet against our $0.04 per 1,000 cannot be computed from published figures. Compare per request, or run both against your own query and count rows.
The most-cited actor on the Apify marketplace, priced per result at $0.40 per 1,000 tweets, with a free plan that cannot call it over the API.
Billing Model
Pay per result on top of an Apify platform plan. The actor page lists $0.40 per 1,000 tweets and quotes throughput of 30 to 80 tweets a second.
$0.40 per 1,000 tweets on the apidojo Tweet Scraper V2 actor
$40 per 100,000 results, the same rate expressed per 100K
Free-plan users run in demo mode only, capped at 5 runs a month and 10 items a run
Free-plan users cannot use the actor via API at all
Cheaper third-party Twitter actors exist on the same marketplace, from $0.15 per 1,000 tweets
Endpoint Surface
Actor runs and datasets rather than REST endpoints, with search, URL, list and profile scraping
Auth Model
An Apify API token in a single header
Free Tier
Demo mode only: 5 runs a month, 10 items each, no API access
Best Fit
Extraction jobs whose orchestration already runs inside Apify
Teams that want a UI-driven actor flow rather than a raw REST call
Multi-platform scraping where Twitter is one source among several
Worst Fit
Real-time work, since actor runs are asynchronous by design
Cost-led bulk pulls, where $0.40 per 1,000 tweets is 10x the TwitterAPIs rate
Anyone hoping to evaluate it free, because API access is not in the free plan
Verdict
Reach for Apify when the workflow shape is worth the premium. For plain tweet extraction it runs an order of magnitude above the cheapest metered option.
Worth knowing that the marketplace is not one price. The headline apidojo actor sits at $0.40 per 1,000 tweets while third-party actors on the same listing page advertise $0.15 per 1,000 tweets and $0.10 per 1,000 follower profiles, so the vendor you actually buy from matters more than the Apify brand on the page.
A general-purpose scraping platform sold as a monthly subscription, whose Twitter endpoint spends 5 credits on every request.
Billing Model
Monthly subscription with a shared credit pool. The Twitter Scraper API costs 5 credits per request, so a plan buys one fifth of the Twitter requests its headline credit count suggests. Failed requests are refunded.
The Twitter Scraper API spends 5 credits per request
LITE at $40 a month covers about 40,000 Twitter requests, an effective $1 per 1,000
STANDARD at $90 a month covers about 200,000 Twitter requests, an effective $0.45 per 1,000
PRO at $200 a month covers about 600,000 Twitter requests, an effective $0.333 per 1,000
The free plan carries 200 credits, which is about 40 Twitter requests
Endpoint Surface
Twitter alongside Google SERP, LinkedIn, Amazon, YouTube, TikTok and a general rotating-proxy scraper, all drawing on the same credit pool
Auth Model
An API key passed as a query parameter or header
Free Tier
200 credits, roughly 40 Twitter requests, enough to smoke-test and no more
Best Fit
Buyers consolidating many scraping targets onto one invoice
Teams that need Google SERP or LinkedIn data beside their Twitter pull
Work that benefits from refunds on failed requests
Worst Fit
Twitter-only workloads, which pay for a platform they do not use
Anyone budgeting off the headline credit count without applying the 5x multiplier
Light or occasional use, where a $40 monthly floor dominates the bill
Verdict
A credible platform buy if Twitter is one of several targets. As a Twitter-only purchase the credit multiplier and the monthly floor both work against it.
Scrapingdog currently holds the top organic result for this query with its own comparison article, which is worth knowing when you read that page: it is a vendor ranking a field it competes in. The credit multiplier is published openly on their pricing page, to their credit, but it is the single most missed number when people size a plan.
The enterprise option: billing per successfully delivered record, unlimited concurrency, SLAs, and a free tier of 5,000 records a month.
Billing Model
Per delivered record, either pay as you go or on a monthly plan, with no charge for failed deliveries. The per-record figure on the pricing panel renders as an animated digit reel rather than static text, so the readable anchor is their own free-tier valuation.
Free tier of 5,000 records a month, no credit card required
Bright Data values those 5,000 monthly credits at about $7.50, which works out to about $1.50 per 1,000 records
The Scale plan includes 384,000 records before additional-record pricing applies
Billing covers only successfully delivered records
Credits renew on the first of each month
Endpoint Surface
A Twitter scraper API within a much larger estate covering proxies, an Unlocker API, a SERP API and dozens of other site scrapers
Auth Model
An API token, with dataset and trigger endpoints
Free Tier
5,000 records a month, renewed monthly, no card
Best Fit
Enterprise procurement that needs an account manager, SSO and a premium SLA
Very high concurrency, which their plans do not cap
Buyers who need compliance posture and vendor scale more than the lowest rate
Worst Fit
Price-led buyers, since the derived rate lands well above every metered option here
Small teams, where the enterprise surface is overhead rather than value
Anyone who needs to read a per-record price off the page without doing arithmetic
Verdict
The right answer when procurement, SLAs and concurrency decide the purchase. It is not the right answer when price per tweet decides it.
The $1.50 per 1,000 records figure above is derived, not quoted: it comes from dividing Bright Data's own stated $7.50 valuation of the 5,000 free monthly credits. We flag that rather than presenting it as a published rate, because their live pricing panel does not render a per-record number in readable text.
Free libraries you run yourself. No licence cost, and no one absorbing the proxies, the account pool, or the next time X changes something.
Billing Model
No software cost. The real bill is residential proxies, X accounts that survive detection, and the engineering hours spent repairing the scraper after each platform change.
$0 in licence fees for the library itself
Residential proxies are the dominant real cost at any serious volume
Maintenance is unbounded, because breakage is triggered by X rather than by your schedule
Running logged in from your own address risks both IP blocks and account suspension
Endpoint Surface
Whatever the specific project implements, commonly search, profiles and timelines
Auth Model
You supply X account cookies or an account pool yourself
Free Tier
Free to install and run
Best Fit
Academic and research pulls where budget is zero and reliability is optional
One-off extractions that do not need to run again next month
Engineers who want full control of the request path and enjoy maintaining it
Worst Fit
Anything on a production schedule, because breakage arrives without warning
Teams without the appetite to manage proxies, CAPTCHAs and banned accounts
Work where legal exposure matters, since driving the X frontend directly carries more of it
Verdict
Genuinely the best answer when the budget is zero and nothing depends on it running tomorrow. It stops being the best answer the moment something does.
Two of the ten results currently ranking for this query are GitHub pages rather than vendors, which tells you how many people arrive here still deciding between buying and building. The honest version of the build case is that the software is free and the operation is not.
The reason these vendors look wildly far apart on price is that almost none of them bill in the same unit. Normalize first, then compare. Here is what each one actually counts.
Provider
Billing unit
Published rate
Cost per tweet comparable?
TwitterAPIs
API call, about 20 tweets
$0.0008 per call
Yes, $0.04 per 1,000 tweets
Apify Tweet Scraper V2
Result, one tweet
$0.40 per 1,000 tweets
Yes, priced per tweet directly
ScrapeBadger
Credit, 1 per request
$0.150 down to $0.054 per 1,000 credits
No, page size not published
Scrapingdog
Credit, 5 per Twitter request
About $1 per 1,000 requests on LITE
No, page size not published
Bright Data
Delivered record
About $1.50 per 1,000 records, derived
Roughly, if a record is one post
Self-hosted open source
Proxy bandwidth and your time
No licence cost
No, cost is operational
Two of the six cannot be reduced to a cost per tweet from published information, and that is worth saying plainly rather than filling the cell with a guess. If you are choosing between a metered provider and a credit provider, the only reliable method is to run the same query on both free tiers and count the rows each one returns for the units it charged you.
What three real workloads would cost
A single headline rate hides the monthly floor. These three profiles price the same work against each billing shape, using each vendor's published rate. Credit vendors are shown per request, since their per-tweet rate is not published.
Scenario 1
Weekend project, sentiment tracker
300,000 tweets a month, which is 15,000 calls at our page size
TwitterAPIs: $12.00 a month
Apify Tweet Scraper V2: $120.00 a month
ScrapeBadger PAYG: $10 minimum buys the credits easily
Scrapingdog: $40 a month minimum plan
Bright Data: about $450 a month at the derived record rate
At this size the monthly floor is the whole story. Metered billing and pay-as-you-go credits both beat any subscription.
Scenario 2
Growth-stage app, steady load
6,000,000 tweets a month, which is 300,000 calls at our page size
TwitterAPIs: $240 a month
Apify Tweet Scraper V2: $2,400 a month
ScrapeBadger Starter: $49 a month for 600,000 requests
Scrapingdog STANDARD: $90 a month for about 200,000 requests
Bright Data: about $9,000 a month at the derived record rate
This is the crossover band, and it turns entirely on page size. A credit plan looks cheaper per request; whether it is cheaper per tweet depends on rows returned.
Scenario 3
Social-listening platform
30,000,000 tweets a month, which is 1,500,000 calls at our page size
TwitterAPIs: $1,200 a month
Apify Tweet Scraper V2: $12,000 a month
ScrapeBadger Growth: $129 a month for 1,800,000 requests
Scrapingdog PRO: $200 a month for about 600,000 requests
Bright Data: enterprise contract territory
At this scale a committed credit plan is clearly cheaper per request than metering. Benchmark the rows returned per request before you move, because that is the variable the plan price does not tell you.
Figures use each vendor's published rate and our own page size of about 20 tweets per call. Real invoices move once you add premium endpoints, write actions or concurrency surcharges. Run your own volume through the cost calculator for a number specific to your workload.
What the market for this term looks like
Search this phrase and the results split three ways, which tells you something about who is actually asking. Roughly half the first page is vendors with a dedicated Twitter scraper product. A quarter is comparison content, some of it published by those same vendors. The rest is GitHub: a topics page and individual scraper repos, read by people still deciding whether to buy anything at all.
That last group is the useful signal. Nobody lands on a query like this having already settled the build-versus-buy question, which is why the open-source route has a card on this page rather than a dismissal. The software genuinely is free. What is not free is the account pool, the residential proxies, and the fact that your maintenance schedule is set by X rather than by you.
Worth noting when you read around: the top organic result for this query is a comparison article published by one of the vendors in this comparison. That is not a criticism, it is the same thing this page is. It just means the ranking you are reading, including ours, is written by a competitor, so weigh the published rates and the free tiers rather than the adjectives. Every number here links to the page it came from for exactly that reason.
Moving from whatever you scrape with today
Most people reading this already collect Twitter data somehow and are weighing a switch. The move is shorter than it looks, because every hosted option here authenticates with a header and returns JSON.
Coming off a self-hosted scraper
The code you delete is larger than the code you write. Proxy rotation, cookie refresh, CAPTCHA handling, account rotation and retry backoff all become one authenticated GET. The part worth keeping is your parsing layer, since you will still map fields into whatever shape your database expects.
Apify returns tweets through a dataset after an actor run, so migrating means trading the trigger-then-poll pattern for direct REST calls. That usually removes the actor scheduling and the polling loop entirely, and the per-tweet cost drops by about 10x at published rates.
Both are a header and a query, so the mechanical work is a base-URL change plus a field-name diff. The number worth recomputing first is your real request volume, since the 5-credit multiplier means your Scrapingdog plan was buying a fifth of the requests its credit count advertised.
Line each X API route up against its equivalent here, then drop the OAuth flow for a single bearer token. The rate-limit scaffolding goes with it: there is no 15-minute window to read, no reset timestamp to sleep against, and no 429 for a quota we set.
For most teams it is TwitterAPIs, at $0.0008 per call returning about 20 tweets, which is $0.04 per 1,000 tweets with no subscription and no platform-imposed rate ceiling. It carries 99 endpoints (63 reads plus 36 writes) and opens with $0.50 in credits and no card. The honest qualifier: at very large committed monthly volume a credit subscription such as ScrapeBadger's Scale plan prices a request lower than a metered call, so "best" depends on whether your volume is steady enough to commit.
On metered pricing with no commitment, TwitterAPIs is cheapest per tweet at $0.04 per 1,000. On price per request at high committed volume, ScrapeBadger is cheapest: its Scale plan works out to $0.054 per 1,000 credits and a Twitter request spends one credit. Those two claims are not in conflict because they measure different units, and a per-tweet comparison against ScrapeBadger is not possible from published figures, since the number of tweets returned per request is not stated. Compare per request, or run a real query on both and count the rows.
No. Every hosted provider on this page runs its own data infrastructure, so you authenticate against the provider rather than against X. On TwitterAPIs that is a single bearer-token header you mint in the dashboard, and you are making live calls within about a minute of signing up. The official X API is the only path that requires an X developer account, app review and an OAuth flow.
Scrapingdog prices each of its endpoints in credits against a shared monthly pool, and the Twitter Scraper API is a 5-credit endpoint, the same tier as its Google SERP Light, LinkedIn Jobs and YouTube scrapers. The practical consequence is that a plan advertising 200,000 credits buys about 40,000 Twitter requests, not 200,000. They publish the multiplier openly on their pricing page, but it is the number people most often miss when sizing a plan against a competitor quoted per request.
Every third-party rate here was read from that vendor's own live pricing or product page on 12 August 2026, and each is attributed to its source in the stats band above. Our own rates come from our published pricing page. Scraping vendors reprice fairly often, and two of the figures here are derived rather than quoted (Bright Data's per-record rate, and every effective per-1,000 rate computed from a plan price divided by its credit allowance), so reconfirm against the vendor before committing budget.
Published rates in 2026 run: TwitterAPIs $0.0008 per call (about $0.04 per 1,000 tweets), ScrapeBadger $0.15 per 1,000 credits pay as you go falling to $0.054 per 1,000 on the $699 Scale plan with 1 credit per request, Apify Tweet Scraper V2 $0.40 per 1,000 tweets, Scrapingdog from $40 a month at 5 credits per Twitter request (about $1 per 1,000 requests on the LITE plan), and Bright Data at roughly $1.50 per 1,000 records derived from their own free-tier valuation. Self-hosted open source has no licence cost but real proxy and maintenance costs.
No provider here ships a free production tier, and the free plans differ more than they look. Apify's free plan runs the Tweet Scraper in demo mode only, capped at 5 runs a month and 10 items each, with no API access at all. Scrapingdog's free plan carries 200 credits, which at 5 credits per Twitter request is about 40 requests. Bright Data gives 5,000 records a month. ScrapeBadger gives 1,000 credits at signup. TwitterAPIs gives $0.50 in credits, roughly 625 calls or about 12,500 tweets, with no card required.
The official X API is first-party and priced accordingly, metering reads from roughly $0.005 per resource, which lands near $5 to $10 per 1,000 tweets. A third-party scraping API runs its own collection layer and resells structured JSON at a fraction of that: TwitterAPIs charges $0.04 per 1,000 tweets on the read tier. The tradeoffs are contractual rather than technical for most teams. If you need a direct commercial relationship with X Corp, use the official API. If you need the data at a workable price, a hosted third-party API returns the same fields.
Only if nothing depends on it. The library is free, and the running cost is not: you supply residential proxies, an account pool that survives detection, and the engineering time to repair the scraper each time X changes its frontend. That repair schedule is set by X, not by you, which is what makes self-hosting hard to run in production. For research pulls and one-off extractions it is a reasonable choice, and two of the ten results currently ranking for this query are GitHub projects rather than vendors.
TwitterAPIs carries 99 endpoints, 63 reads plus 36 writes, spanning search, tweet detail, replies, retweeters, thread expansion, profiles, timelines, mentions, the follower graph, list members and DM reads, with write actions covering likes, reposts, bookmarks, follows, tweet creation and media upload. Most scraping-first vendors expose a narrower read-only slice built around search and profiles, which is fine for tweet collection and a wall the moment you need the follower graph, a full thread, or the ability to act rather than only observe.