The best X (Twitter) scraping APIs of 2026
APIs that return public X posts, profiles and followers, including the official X API.
Researched by Which Social API editorial teamLast reviewed 29 October 2026How we review
Which Social API uses a mix of qualitative and quantitative review methods, based on each provider’s documentation, published pricing and public user reviews. We never accept money in exchange for a higher ranking or a better review. Who runs this site
X ended its free API in 2023, and in 2026 it replaced its subscriptions with pay-per-use. Reading a post officially now costs $0.005 and a follower record $0.010, with a cap of 3 million post reads a month before you’re into an Enterprise contract. Third-party APIs sell the same public posts for $0.15 to $0.20 per 1,000.
That 25–33× gap is why this market exists, and our ranking follows it. For most developers, twitterapi.io is the best X data API: the cheapest, the most endpoints, and real-time pushes. SocialData is the cleaner, read-only alternative. The official API ranks third. It’s the only sanctioned source and the one to use if you post or your legal team needs it, but on price it loses every comparison.
The multi-platform APIs trail. Most cover X thinly: Bright Data has no search, and ScrapeCreators returns only a user’s most popular tweets.
Our picks
Full ranking
- 1twitterapi.ioMost developers reading X at volume8.5
- 2SocialDataSimplest read-only X API8.1
- 3X APISanctioned access and full-archive search7.3
- 4ApifyScheduled X exports without code7.0
- 5Data365Scheduled X collection for listening teams6.5
- 6TikHubPay-per-request X calls alongside Asian platforms6.2
- 7Bright DataBulk posts from a known list of accounts6.0
- 8ScrapeCreatorsOccasional X lookups on a multi-platform key5.4
Before you choose
On X, the price per 1,000 posts gets you started, but the edge cases decide the bill.
- Count the empty calls. twitterapi.io charges at least 15 credits for every call, and SocialData charges for empty requests past three a minute. Keyword polling that finds nothing still costs money.
- Check the X endpoints, not the logo. “Covers X” can mean seven endpoints. Bright Data has no search, ScrapeCreators returns about 100 popular tweets per user, and Apify’s most-used Actor won’t fetch replies.
- Followers are where prices split. A follower record costs $0.010 on the official API, $0.0002 on SocialData and as little as $0.00001 on twitterapi.io in bulk. For a 100,000-follower account that’s $1,000, $20 or $1.
- Know who you’re paying. twitterapi.io is a small Delaware LLC, SocialData names no company at all, and TikHub’s terms let it sell data derived from customer requests. Top up small amounts and read the refund terms.
- Keep writes off accounts you care about. Third-party write endpoints run on your own logged-in X account. If posting matters, the official API is the safer route.
Four ways to get X data
Which one fits depends on how much you read, whether you need to post, and how much access risk you’ll accept.
Official X API
X’s own API, billed per resource returned.
- Good for
- Sanctioned access, full archive, writes
- Watch for
- $5 per 1,000 posts, 3M monthly cap, frequent changes
- Examples
- X API
X specialists
Unofficial APIs that do nothing but X, priced per item.
- Good for
- Lowest prices, deepest X coverage, monitors
- Watch for
- Small operators, no guarantee of continued access
- Examples
- twitterapi.io, SocialData
Multi-platform APIs
One key for X and many other networks.
- Good for
- Teams already using them elsewhere
- Watch for
- Thin X endpoints, higher prices per post
- Examples
- Data365, TikHub, ScrapeCreators, Bright Data
Scraper marketplaces
Third-party scrapers run as jobs on a hosted platform.
- Good for
- Scheduled exports without code
- Watch for
- Quality and rules vary by author
- Examples
- Apify
Prices and coverage
Published prices and what each one returns, side by side. How we review
| Provider | Per 1,000 posts | Plans from | Free trial | Billing |
|---|---|---|---|---|
| twitterapi.io | $0.15 | Pay as you go, or $29 a month | $1 of credit, no card | Prepaid credits, per item |
| SocialData | $0.20 | Pay as you go | Unclear; demo account on request | Prepaid balance, per item |
| X API | $5.00 | Pay per use | None | Prepaid credits, per resource |
| Apify | $0.25 | Free ($5 credit), then $19 a month | $5 of credit a month | Monthly plan, per result |
| Data365 | Subscription | €300 a month (Basic, one network) | 14 days, on request | Monthly credits |
| TikHub | $1.00 | $0.001 a request, $5 minimum top-up | $0.05 of credit | Prepaid, per request; all sales final |
| Bright Data | $1.50 | $1.50 per 1,000 records, pay as you go | 5,000 records a month free | Per record |
| ScrapeCreators | $1.88 | $47 for 25,000 credits | 100 free credits | Prepaid credits that never expire |
| Provider | Profiles | Posts | Replies | Followers | Search | Streaming |
|---|---|---|---|---|---|---|
| twitterapi.io | Yes | Yes | Yes | Yes | Yes | Yes |
| SocialData | Yes | Yes | Yes | Yes | Yes | Yes |
| X API | Yes | Yes | Yes | Yes | Yes | Yes |
| Apify | Yes | Yes | No | Yes | Yes | No |
| Data365 | Yes | Yes | Yes | Yes | Yes | No |
| TikHub | Yes | Yes | Yes | Yes | Yes | No |
| Bright Data | Yes | Yes | No | No | No | No |
| ScrapeCreators | Yes | Yes | No | No | Yes | No |
Green marks the best result in each column.
The providers in detail
1. twitterapi.io
Most developers reading X at volumeVisit twitterapi.ioout of 10
- Data
- Tweets, search, replies, profiles, followers, lists, communities, Spaces, trends, writes
- Freshness
- Live
- Plans from
- Pay as you go, or $29 a month
- Free trial
- $1 of credit, no card
- Billing
- Prepaid credits, per item
- SDKs
- REST only, MCP server
twitterapi.io is the cheapest serious way to read X. Tweets cost $0.15 per 1,000 on pay-as-you-go, against $5.00 on the official API, and bulk follower pages drop to $0.01 per 1,000 users. It also has the longest endpoint list of the X specialists: about 75, covering search with X’s operators, replies, quotes, followers, lists, communities, Spaces and trends, plus webhook and WebSocket pushes for keyword rules and monitored accounts.
You start with $1 of free credit and no card, but the account is held to one request every five seconds until you pay. After that it’s 20 requests a second, rising to 500 on the $1,499 Business plan. There’s no SDK, just plain HTTP, and over-limit calls return a 429 with no Retry-After header. Subscriptions from $29 a month add bonus credits; the vendor’s “as low as” rates assume you use every bonus credit before it expires.
The doubts are about the company, not the price. It’s run by Prism Digital, LLC in Delaware, the domain dates from December 2024, and no team or funding is published. Its terms say it provides no data that requires authentication, yet it sells login, DM and bookmark endpoints that act on your own X account. Use it for reads, keep the write endpoints away from any account you care about, and remember that keyword polling pays for empty results too.
Good
- $0.15 per 1,000 tweets, and $0.01 per 1,000 followers in bulk
- The widest endpoint list of the X specialists, including trends and real-time pushes
- $1 of free credit with no card
Not so good
- Small operator with thin company details
- Its terms rule out data that needs a login, yet it sells login, DM and bookmark endpoints
- Every call costs at least 15 credits, even when nothing comes back
3. X API
Sanctioned access and full-archive searchVisit X APIout of 10
- Data
- Posts, full-archive search, users, followers, lists, Spaces, Communities, Notes, DMs, trends, writes
- Freshness
- Live, first-party
- Plans from
- Pay per use
- Free trial
- None
- Billing
- Prepaid credits, per resource
- SDKs
- XDK, MCP server
The official API is the only way to get X data with no question over access. Pay-per-use includes full-archive search back to 2006, a filtered stream, DMs, writes, Community Notes, an official SDK (XDK) and an MCP server. If you need to post, or your legal team won’t sign off on anything else, this is your pick.
It’s also the most expensive by a distance. Reads cost $0.005 per post and $0.010 per user or follower record, so a 100,000-account follower list costs $1,000, against about $1 at twitterapi.io’s bulk rate. Pay-per-use is capped at 3 million post reads a month, about $15,000, and past that you’re negotiating Enterprise. From 22 November 2026, expansions (the author and media objects bundled with posts) are billed too.
The bigger problem is trust in the bill. In 2026 X moved Basic and Pro subscribers onto pay-per-use with $200 and $5,000 initial credit charges, repriced posts containing links to $0.20 each, and restricted programmatic replies. Its developer forum carries reports of 402 errors on positive balances and missing vouchers. Use it where you must; use a third party where you can.
Good
- The only sanctioned source, with full-archive search back to 2006 on pay-per-use
- Writes, DMs, a filtered stream and an official SDK
- No subscription or minimum spend
Not so good
- $5 per 1,000 posts and $10 per 1,000 followers, 25–33× third-party prices
- Capped at 3 million post reads a month before Enterprise
- Frequent price and policy changes, and billing complaints on X’s own forum
4. Apify
Scheduled X exports without codeVisit Apifyout of 10
- Data
- Tweets by search, profile, hashtag and date; profiles; followers and following
- Freshness
- Live, per run
- Plans from
- Free ($5 credit), then $19 a month
- Free trial
- $5 of credit a month
- Billing
- Monthly plan, per result
- SDKs
- JavaScript, Python, CLI, MCP server
Apify doesn’t build its own X scraper. Every popular X Actor is third party, so you’re choosing an author as much as a platform. The most used, API Dojo’s Tweet Scraper V2, costs $0.40 per 1,000 tweets but needs at least 50 tweets per query, won’t fetch single tweets or reply threads, and runs one job at a time. It holds a 3.86 rating from 196 reviews. kaitoeasyapi’s tweet scraper is cheaper, at $0.22 per 1,000 on Starter and $0.18 on Business. Its author name matches the handle twitterapi.io launched under on Hacker News; we haven’t confirmed they’re the same operator.
What you pay the overhead for is the platform: scheduling, webhooks, datasets you can export as CSV or Excel, Make and Zapier integrations, and JavaScript and Python clients. That makes it the best option here for someone who wants a weekly export of a search without writing a client. Tweet Scraper V2 runs in demo mode on the free plan, so budget for Starter at $19 a month.
For a product that pulls X data on demand, a direct API is simpler. Actors run as jobs rather than single calls, and each one has its own input rules and output shape.
Good
- Cheap per tweet on third-party Actors, from $0.18 per 1,000
- Scheduling, webhooks, CSV and Excel exports, and no-code integrations
- The same account covers every other platform
Not so good
- No Apify-built X Actor, so quality depends on the author
- Tweet Scraper V2 needs 50 tweets per query, skips replies and runs one job at a time
- Each Actor has its own rules, schema and maintainer
5. Data365
Scheduled X collection for listening teamsVisit Data365out of 10
- Data
- Profiles with posts, replies, retweets, quotes, followers, following, post and profile search
- Freshness
- As fresh as your last update task
- Plans from
- €300 a month (Basic, one network)
- Free trial
- 14 days, on request
- Billing
- Monthly credits
- SDKs
- None (Postman, Python and JS examples)
Data365 covers X well for a multi-platform API: profiles with posts, followers and following, replies, retweets, quotes, post search and profile search. An X post costs 1 credit, which on the €300-a-month Basic plan works out to €0.60 per 1,000. Profiles cost 9 credits, about €5.40 per 1,000.
It’s built for monitoring teams rather than apps. Every request is a task: you POST to start a collection, wait or take a callback, then GET the result from Data365’s store. The vendor puts that at one to five minutes a task. Auto-update tasks re-run collections on a schedule, which suits social listening.
Against twitterapi.io or SocialData it’s slower, needs a €300 monthly commitment for a single network (€850 for three), and has no SDK or status page. Reviews are positive but few: 12 on Trustpilot. Pick it if X is one of several networks you track on a schedule.
Good
- Followers, replies, quotes and search on X
- Auto-update tasks re-run collections on a schedule
- 14-day trial and a named support manager
Not so good
- €300 a month minimum, for one network
- Asynchronous tasks the vendor puts at one to five minutes each
- No SDK and no public status page
6. TikHub
Pay-per-request X calls alongside Asian platformsVisit TikHubout of 10
- Data
- Tweets, search, trending, replies, profiles, followers, retweeters, lists, communities
- Freshness
- Live
- Plans from
- $0.001 a request, $5 minimum top-up
- Free trial
- $0.05 of credit
- Billing
- Prepaid, per request; all sales final
- SDKs
- Python, Java, TypeScript, Go, C++
TikHub’s X coverage is decent for a generalist: 25 endpoints, including search, trending, followers and followings, retweeters, comments, lists and communities. At $0.001 a request, falling to $0.0005 past 30,000 requests a day, with a $5 minimum top-up, it’s a cheap way to add X calls if you already use it for TikTok or the Chinese platforms. Some endpoints cost up to $0.01 a request, and you only see per-endpoint prices after signing up.
The terms are what hold it back. Section 8.5 lets TikHub sell aggregated, de-identified data derived from the public content its API accesses for customers; you can opt out by email. All sales are final, with refunds only at its discretion. Its status page returned an error (HTTP 521) when we checked on 29 October 2026, and there are only two reviews on Trustpilot.
For X alone, twitterapi.io and SocialData go deeper and sell on cleaner terms.
Good
- Pay per request from $0.001, with a $5 minimum top-up
- Search, followers, lists and communities on X
- SDKs in five languages
Not so good
- Its terms let it sell data derived from what customers’ requests collect
- All sales final; refunds only at TikHub’s discretion
- Its status page was down when we checked
7. Bright Data
Bulk posts from a known list of accountsVisit Bright Dataout of 10
- Data
- Posts by URL or profile, with views and bookmarks; profiles
- Freshness
- Live; datasets refreshed monthly
- Plans from
- $1.50 per 1,000 records, pay as you go
- Free trial
- 5,000 records a month free
- Billing
- Per record
- SDKs
- Python, MCP server
Bright Data is the company X sued and failed to stop: a federal judge dismissed X’s claims over scraping and selling public posts in May 2024, and the narrower claims that survived ended in a confidential settlement in June 2025. That history is reassuring if your lawyers ask. Its X product is less so. There are two scrapers, posts and profiles, collected by URL or discovered from a profile URL with a date range.
There’s no keyword or advanced search and no follower lists, which rules it out for most X jobs. What it does return is rich: post records include views, likes, reposts, quotes, bookmarks and the quoted or parent post. Pricing is $1.50 per 1,000 records on pay-as-you-go and $1.30 on the $499 Scale plan, with 5,000 free records a month shared across its products.
Use it to bulk-collect posts from accounts you already know, ideally if you already buy from Bright Data. For anything that starts with a search, look higher up this list.
Good
- Rich post records, including views, bookmarks and the quoted or parent post
- Bulk asynchronous delivery to a webhook or S3
- The strongest legal track record against X of anyone here
Not so good
- No keyword or advanced search on X
- No follower lists or reply threads
- Records that fail on bad input are still billed
8. ScrapeCreators
Occasional X lookups on a multi-platform keyVisit ScrapeCreatorsout of 10
- Data
- Profiles, about 100 popular tweets per user, tweet details, transcripts, search, communities
- Freshness
- Live, optional cache
- Plans from
- $47 for 25,000 credits
- Free trial
- 100 free credits
- Billing
- Prepaid credits that never expire
- SDKs
- REST, CLI, MCP server
ScrapeCreators is strong on TikTok and Instagram, but X is an afterthought. It has seven X endpoints: profile, user tweets, tweet details, tweet transcript, search, community and community tweets. By the vendor’s own docs, user tweets returns only about 100 of a user’s most popular tweets, not the latest. There are no follower lists and no full timeline.
It’s still pleasant to use. One key covers nearly 40 sources, credits never expire, and most requests cost 1 credit ($1.88 per 1,000 on the $47 pack, $0.99 on the $497 pack). The docs and changelog are better than most. If you already pay for it for other platforms, a profile lookup or a search on X needs no new account.
Don’t buy it for X. Anyone tracking accounts over time or pulling conversations needs twitterapi.io or SocialData.
Good
- One key and never-expiring credits across nearly 40 sources
- Search and tweet transcripts
- Good docs and a detailed public changelog
Not so good
- User tweets returns about 100 popular tweets, not the latest
- No followers, replies or full timelines
What X data you can get
Public posts and profiles are easy. Past that, coverage varies a lot by provider, and some data is off-limits to everyone.
- Profiles
- Everyone
- Bio, counts and verification. The official API charges $0.010 a user; third-party bulk lookups cost far less.
- Posts and timelines
- Everyone, with limits
- ScrapeCreators returns about 100 popular tweets per user. Bright Data needs a post or profile URL.
- Replies and threads
- Most
- Not on Bright Data, ScrapeCreators or Apify’s most-used Actor.
- Search
- Most
- X search operators on twitterapi.io and SocialData. Official full-archive search reaches back to 2006 on pay-per-use. Bright Data has none.
- Followers and following
- Most
- $10 per 1,000 officially, $0.20 on SocialData, from $0.01 on twitterapi.io in bulk. Not on Bright Data or ScrapeCreators.
- Real-time
- Few
- The official filtered stream allows one connection and 1,000 rules on pay-per-use. twitterapi.io and SocialData sell per-account monitors.
- View counts
- Varies
- Shown on the site, and Bright Data lists views in its post fields.
- Likes
- Private
- X made everyone’s likes private in June 2024.
- Protected accounts, DMs, deleted posts
- Not available
- DMs only for your own account, through the official API.
Is it legal to use a third-party X API?
X says no. Its terms ban crawling or scraping without written consent and set liquidated damages of $15,000 per million posts accessed in a day, extended to anyone who knowingly helps someone else break them. Its developer terms steer commercial use beyond self-serve to Enterprise.
Courts have been less sure. In X Corp v. Bright Data, a federal judge dismissed X’s claims over scraping and selling public posts in May 2024, holding them preempted by copyright law. Narrower claims about harm to X’s servers went on until a confidential settlement in June 2025. That case was about logged-out access, and X has required a login to view posts since mid-2023. Anything that runs on logged-in accounts, including third-party write endpoints, is on weaker ground.
Our view: reading public posts through a third party is a contract and business risk to weigh, not a settled question. If you’re bound by X’s developer terms, or can’t afford a dispute, use the official API. In the EU and UK, GDPR treats handles, bios and posts as personal data, whoever supplies them.
This is a summary, not legal advice.
Why X requests fail
We haven’t measured failure rates yet. These are the problems that come up most in public issue trackers and developer forums.
Login walls and account pools
Since mid-2023 most X content needs a logged-in session. Unofficial tools rely on pools of real accounts, which X locks, rate-limits or challenges.
Changed internal endpoints
X changes its internal endpoints and response shapes. Open-source scrapers broke in April 2026, and others dropped retweets when the author object came back incomplete.
Empty or failed?
Some tools can’t tell a failed request from an empty result. SocialData and twitterapi.io both charge for some empty results, so monitoring jobs pay for “nothing found”.
Rate limits
The official API counts in 15-minute windows. twitterapi.io allows one request every five seconds until you pay, and returns 429s without a Retry-After header. SocialData defaults to 120 requests a minute.
Upstream outages
When X goes down, every third-party API goes down with it. twitterapi.io users have tied brief outages to X’s own.
Billing errors on the official API
Developers report 402 “credits depleted” errors on positive balances. Auto-recharge runs at most once every five minutes, and a spend cap blocks requests until the next cycle.
What a request looks like
The official recent-search endpoint covers the last seven days. Third-party APIs take a similar search query with their own API key and charge a fraction of the price.
# Official X API, pay-per-use
GET /2/tweets/search/recent?query=<your search>
Authorization: Bearer <app-only token>
# Up to 100 posts per request, billed at $0.005 per post returned.
# Limit: 450 requests per 15 minutes per app.Questions
What’s the cheapest X (Twitter) API?
twitterapi.io, at $0.15 per 1,000 tweets on pay-as-you-go. SocialData charges $0.20. The official API charges $5.00.
Is the official X API free?
No. The free tier ended when pay-per-use launched in February 2026. There’s no minimum spend, but every read is billed, and only apps X classes as public utilities get free scaled access.
Do I need an X developer account?
Not for the third-party APIs here: they use their own API keys. The official API needs an app in X’s developer console, and some apps get “not permitted” errors until X reviews the use case.
Can I search old tweets?
Yes, through the official API: full-archive search back to 2006 is available on pay-per-use, despite pages that say it’s Enterprise-only. twitterapi.io and SocialData support X’s search operators too.
Can I post through these APIs?
Through the official API, and through twitterapi.io, which posts from your own X account. SocialData is read-only. On the official API, a post containing a link costs $0.20.
Is it legal to use a third-party X API?
X’s terms forbid scraping, but a federal court dismissed X’s main scraping claims against Bright Data in 2024. It’s a contract risk to weigh. See the section on the law above.
How often is this ranking updated?
We review it regularly, and sooner when a provider changes its pricing or features or users report a problem.
Can providers pay to rank higher?
No. We never accept money in exchange for a higher ranking or a better review.
Quoting this ranking?
You’re welcome to. Please link back to this page.
Which Social API (2026). The best X (Twitter) scraping APIs of 2026. whichsocialapi.com/best/x-scraping-apis
2. SocialData
Simplest read-only X APIVisit SocialData ↗out of 10
SocialData does one job and keeps it simple: public X data at a flat $0.0002 per tweet or profile, so $0.20 per 1,000, with no plans or seats. Failed requests aren’t charged, the balance doesn’t expire, and an unused deposit is refundable within 14 days. It costs a third more than twitterapi.io per tweet, and far more for big follower lists ($0.20 per 1,000 against $0.01), but the bill is easier to predict.
The docs are the cleanest here: every endpoint page shows its price, and there’s an OpenAPI spec and an MCP server. The Monitoring API pushes signed webhooks when an account posts, follows or edits its profile, at about $4.99 an account a month for the first ten. Search monitors are the trap. At the default 30-second refresh, one costs about $17.28 a month before it finds anything.
It’s read-only, which we count as a plus if you don’t want account risk. The minuses: 120 requests a minute by default (more is free on request), email-only support, and no named company behind it. Pick it over twitterapi.io if you want predictable reads and don’t need writes or trends.
Good
Not so good