Skip to content

Twitter and X API answers

Quick answer

TwitterAPIs is a pay-per-call Twitter and X API that returns tweets, profiles, search results, and direct messages as JSON without an X developer account. This page indexes our direct answers to the cost and access questions developers search before they build, and every entry below opens with the number rather than a preamble. Standard reads bill at $0.0008 per call, per our own published pricing, across 99 public endpoints, 63 read and 36 write, of which 60 carry that rate, 24 are free and 15 sit between $0.0016 and $0.01.

TwitterAPIs is a Twitter and X API alternative, and this page is its direct-answer index for the questions developers actually search before they touch the API: what it costs, how the pricing tiers work, whether a free option exists, and how the official API compares to third-party alternatives. Each entry below opens with the number itself, not a preamble, and links out to the full pricing or setup detail for whoever needs the longer version. One note on how to read the numbers: what you are billed is $0.0008 per call, whatever that call returns. Every "per 1,000 tweets" figure below, ours and everyone else's, is derived from a per-call rate by assuming a full 20-tweet page, which is the default page size rather than a guaranteed yield.

The four questions people arrive with

Most visits start with one of these. Each is answered here in a sentence, with the full version a click away, so you do not have to open a page to find out whether it is the one you wanted.

How much does the X API cost?

Here, $0.0008 per call, whatever that call returns, with no monthly minimum. The official API is sold in monthly tiers instead, which is why the two are hard to compare directly: one bills what you use and the other bills what you reserved. The comparison that actually decides it is not the headline rate but your duty cycle. A job that runs for an hour a day and sits idle the rest of the time pays for one hour here and a full month on a tier, so the same workload can look cheap or expensive depending only on how evenly it is spread. Work out your monthly call volume first, then multiply, then compare that against the tier you would otherwise have to buy. Note also that per-call means per REQUEST, not per record returned, so a request that comes back empty costs exactly what a full page costs. That makes tight filters cheaper in result terms and more expensive per useful row, which is the opposite of the intuition most people bring from per-record pricing.

The full cost breakdown

Is there a free X API tier in 2026?

Not a useful one for reading data. What exists here instead is free credit on signup with no card, which is enough to test whether the endpoints return what you need before you spend anything. Treat that credit as a correctness check rather than a trial period: the question it answers is whether the response shape matches what your code expects, and that is answered in a handful of calls rather than over a month. The thing worth checking during it is the field set, because that is what silently breaks integrations later. Pull one record from each endpoint you intend to use, print the actual keys, and write your parser against those rather than against a documentation example. A free tier that lets you confirm the data is right is worth more than a larger one that only lets you confirm the service responds.

What free actually gets you

How do you get an X API key?

Two different things share that name: a key for the official developer platform, which needs an approved project, and a key here, which is a signup. Which one you need follows from what you are doing rather than from the name. Anything that writes as a specific user, posts, or touches a protected account needs the official platform, because that authorisation is granted by the platform and cannot be delegated. Reading public data does not, which is the whole reason a second kind of key exists. The practical difference is the wait: one is an application with an approval step and a stated use case, the other is available immediately. If you are prototyping and do not yet know which side of that line your project falls on, start with the one that does not block, and move only the parts that genuinely require official write access.

Which one you need

When is the official API worth the cost?

When you need write access at scale, official rate-limit guarantees, or a compliance story that names the platform itself. Reading public data at volume is where the tier pricing stops making sense, and that is the case migrating off it usually rests on. Those three reasons are worth taking seriously rather than arguing with. If a contract or a review requires that your data comes from the platform under its own terms, no price comparison changes that, and a cheaper route is the wrong answer however much cheaper it is. The honest version of the migration question is therefore narrower than it sounds: it applies to the read-heavy, public-data part of a workload, which is usually most of the volume and rarely the part anyone is contractually attached to. Splitting a system that way is often better than moving all of it, because the writes stay where they have to be and the reads stop being priced against a tier that was sized for them.

The comparison, including where the official API wins

Pricing, plainly

Every figure here is interpolated from the same constants the API bills on, so this paragraph cannot drift from the product the way a typed number would. Standard reads are $0.0008 per call, charged per call rather than per record, so a call that returns nothing costs the same as one that returns a full page. There is no monthly minimum and no seat cost.

Across 99 public endpoints, 24 bill nothing at all, 60 carry the standard read rate, and 15 sit above it, between $0.0016 and $0.01, because they cost materially more to serve. The split matters more than the headline rate: which bucket your workload lands in decides your bill, not the average.

Cost, pricing tiers and the free question

What a call costs, how the tiers are drawn, and how the bill compares with the official X API. Standard reads bill at $0.0008 per call and 24 of our 99 endpoints bill nothing at all.

Access, keys and authentication

How to get a key, what that key is allowed to do, and how to reach X data without applying for an X developer account first.

Profiles, the follow graph and saved items

Turning a handle into the numeric id every other endpoint wants, searching people rather than posts, reading who follows whom, and exporting likes and bookmarks.

Posting, direct messages and media

The write half of the surface: publishing posts and Articles, sending DMs, deleting in bulk, and pulling the video file out of a post. 36 of the 99 endpoints write rather than read.

Realtime, monitoring and webhooks

Watching accounts and keywords while they move: polling against webhooks, following one post over time, trends, Spaces, Communities and Grok.

Rate limits, pagination and the legal position

What rate limiting means on X, how paging through a large result set works, whether proxies are part of the job, where collecting public posts stands legally, and what the endpoint catalogue holds.