How-To Guide
How much does the X (Twitter) API cost?
The X API bills per call: $0.005 a post read, $0.015 a plain post, $0.200 a link post and $0.001 an owned read, with a 3 million monthly cap on reads.
By the Superdirector editorial team. X's official developer documentation (the pricing, free-credits and Enterprise pricing pages at docs.x.com) and every page cited here were read on November 14, 2026; vendor prices are as each page prints them on that date.
At a glance
How much does the X (Twitter) API cost?
Per call, and nothing by subscription. Reads bill per returned resource: $0.005 a post, $0.010 a user, $0.001 a like, and $0.001 for reading your own data. Writes bill per request: $0.015 a plain post, $0.200 a post that carries a link, $0.010 a summoned reply. Credits loaded in the Developer Console cover it, pay-per-use plans cap at 3 million post reads per billing cycle, and above that Enterprise is a custom contract. There is no free tier for reads; new paying accounts earn up to $70 of one-time starter credits that expire after 3 months.
- Every rate on this page was read on X's own developer documentation on November 14, 2026; the Developer Console always carries the current card.
- The two prices that surprise people: $0.200 for a post containing a URL against $0.015 for a plain post, and $0.001 per resource for reading your own data.
- The volume ceiling: 3 million post reads per monthly billing cycle, with Enterprise above it as a custom contract with no post read cap.
- No free tier for reads. Up to $70 of one-time starter credits delays a new account's first invoice; they expire after 3 months.
A post carrying a URL prices at $0.200 per request against $0.015 for a plain post, 13.3 times as much by arithmetic, and the bars share one scale. Original diagram, hand-drawn from X's pricing page read November 14, 2026.
Make it work for your brand
Find it. Remake it. Post it. Superdirector turns the viral formats in your niche into your brand's carousels, X posts and video scripts, and keeps the calendar full.
The official rate card, per resource and per request
Start with the page X itself maintains, because everything else on the internet restates it. X's official pricing documentation opens on the billing model in one line: the X API uses pay-per-usage pricing, no subscriptions, pay only for what you use. You purchase credits upfront in the Developer Console, credits are deducted as you make API requests, and different endpoints carry different costs. All prices are measured per resource fetched for reads, or per request for writes and actions. The only caveat the page stamps everywhere is that prices are subject to change and the Developer Console always carries the current card, so treat the table below as the card as we read it on November 14, 2026.
| Operation | Official unit cost (read on 2026-11-14) |
|---|---|
| Post: read | $0.005 per resource returned |
| User: read | $0.010 per resource returned |
| DM event: read | $0.010 per resource returned |
| Following/followers: read | $0.010 per resource returned |
| Like: read | $0.001 per resource returned |
| Owned read (your own data) | $0.001 per resource returned |
| Post: create (plain) | $0.015 per request |
| Post: create (contains a URL) | $0.200 per request |
| Post: create (summoned reply) | $0.010 per request |
| DM or user interaction: create | $0.015 per request |
| Interaction: delete | $0.010 per request |
| Post counts: recent / all | $0.005 / $0.010 per request |
One row deserves its own paragraph because it is the line that makes publishing workflows affordable: Owned Reads. These are requests made by your own developer app for your own data, from your own posts to your mentions, bookmarks and followers, and they price at $0.001 per resource, 1,000 resources for $1. The discount has a precise boundary: the {id} in the endpoint must match the authenticated user, and that user must be the owner of the developer app. Your own timeline is cheap to read; other people's accounts never qualify.
What moves the bill: links, batch reads and the dedup window
Three mechanics decide whether a month costs a few dollars or a few hundred. The first is the link surcharge. The write card prices a plain post at $0.015 and a post containing a URL at $0.200, more than thirteen times the plain rate by arithmetic. Run the two rows against a real week of publishing: 100 plain posts bill $1.50, while 100 link posts bill $20 from the same card, on the same numbers, with no other charge added.
The second mechanic is per-resource billing on batches. Reads are charged per resource returned in the response, so one search call that returns 100 posts bills as 100 resources at $0.005 each, not as one $0.005 call. Blotato's guide on the same card says it directly: per-resource pricing means batched requests still cost real money. That is the row that catches budget planners who assumed price per call rather than per item.
The third mechanic is the dedup window, and it only softens the other two. All resources are deduplicated within a 24-hour UTC day window, the window resets at midnight UTC, and requesting the same post twice in that window counts as one charge. X calls the mechanism a soft guarantee: edge cases like service outages can result in resources not being deduplicated, so treat dedup as a discount you usually get rather than a line item you can forecast against.
- Set a spending limit before you scale anything. When the limit is reached, API requests are blocked until the next billing cycle, which turns an accident into a pause rather than an invoice.
- Add credits before the balance reaches zero. The docs warn that a balance can go slightly negative and requests stay blocked until credits cover the negative part.
- If you turn on auto-recharge, know its three rules: it requires a saved default payment method, it can trigger at most once per 5-minute window, and it stays paused while the balance is zero or negative.
The two ceilings: 3 million post reads and Enterprise
Pay-per-use is not unlimited. The official pricing page caps such plans at 3 million post reads per monthly billing cycle and routes anything higher to an Enterprise plan. Enterprise is not a checkout option: its own pricing page describes plans as custom-tailored to your organization's needs, with pricing tied to data volume and throughput under committed contracts, no post read cap, volume and likes streams (the full firehose) that pay-per-use cannot buy, and a named account manager with priority support included on every plan.
The flat tiers some guides still quote are gone by their own account. The official pages we read on November 14, 2026 list only pay-per-use and Enterprise; there is no Basic or Pro row to price against anymore. On the history, we print what vendors report rather than guessing: Blotato's article says legacy Basic subscribers were force-migrated to pay-per-use after June 1, 2026, and that X announced on August 14, 2026 that legacy Pro ($5,000 a month) is deprecated too, with remaining subscribers migrating automatically at their cutover dates. Their page carefully adds that the old allowances are the tiers' last published figures, not a current card, and that is the honest frame for them everywhere: retired history, no longer purchasable, quoted only as third-party memory.
Free the way the docs mean it: starter credits, not a free tier
So is there a free X API? X's own pages answer with credits, not with a free tier. A first eligible payment card saved to a new account earns $20 in free credits once the account completes its first credit purchase, and turning on auto-recharge earns a one-time match of that first recharge up to $50. A new account that earns both receives $70 in free credits in total. Three rules bound them sharply: free credits expire 3 months after they are granted, they are spent before your purchased credits, and the card incentive changed on October 13, 2026 so that adding a card on its own no longer qualifies without a first purchase.
- Save a first eligible card on a new account, then complete a first purchase of API credits: this grants the one-time $20. Prepaid cards do not qualify, and adding a card on its own stopped qualifying on October 13, 2026.
- Turn on auto-recharge with a saved default payment method: the first automatic recharge triggers a one-time matching credit of up to $50.
- Note the spending order and clock: free credits are spent before purchased credits and expire 3 months after they are granted.
Beyond the starter credits there is one small earn-back program, and it does not discount the X meter. Purchasing X API credits can earn free xAI API credits based on cumulative spend during a billing cycle, from 10% over $200 up to 20% at $1,000 or more, and collecting it requires linking your xAI team to the X developer account. Those credits spend on xAI's models, not on X's meter, so they reward teams already building Grok features and do nothing for someone counting posting costs.
The buyer-facing verdict on reads stays blunt. A third-party comparison page we cross-checked, stamped October 1, 2026, answers the free question with: not for reads, every request draws from a prepaid balance. The cheapest honest route for a publishing workflow is the owned-read row at $0.001 per resource, reading only your own account's data; the cheapest thing X hands a brand-new paying account is a $70 countdown that expires.
Two one-time starter credits make $70: $20 once a new account saves a first eligible card and completes a first credit purchase, then a match of the first auto-recharge up to $50. Both expire three months after granting and are spent before purchased credits. Original diagram, hand-drawn from X's free-credits page read November 14, 2026.
Pay the meter, rent a wrapper, or buy operated posting
The card prices operations, so price your operation. A month of 300 plain posts bills $4.50 and 720 owned reads bill $0.72, so a light own-account publisher lands under $10 on the meter. The same arithmetic turns cruel at link volume or read scale: 100 link posts cost $20, and a monitor reading 10,000 posts and 10,000 profiles bills $50 plus $100 by the official rows. The table below sorts the common cases.
| Your job | What the meter does to it | The row that changes the answer |
|---|---|---|
| Publish a few plain posts a day from your own account | A few dollars a month, mostly on owned reads | Owned Reads at $0.001 per resource |
| Auto-publish links (newsletters, product pages) | $0.200 per post, 13.3 times a plain post by arithmetic | Post: Create (with URL) $0.200 per request |
| Monitor mentions and other people's accounts | Bills fast as resources add up, and stops at 3 million post reads a month | 3 million Post reads per monthly billing cycle |
| Read your own data for dashboards | The cheap side of the card | Owned Reads $0.001 per resource |
| You just want posts to happen on schedule | A meter to build, feed and watch | The alternatives below |
Three ways remove the meter. Build direct is right when the API's own features are your product: monitoring tools, analytics, research on the full data surface, in which case per-resource billing is the price of the data. Rent a wrapper when publishing is all you need: Blotato's own page prices its API on paid plans from $29 a month for Starter, one subscription whose publishing calls carry no per-post fees of the meter's kind. Buy operated posting when you want the calendar handled, not just the pipe: Superdirector (ours) does that at $199 a month for one AI-run operated social account, with posting permission standing at the account level, while its $29 a month tier is MCP access plus hands-on creation, the version you run yourself. The honest summary: the meter is a fair price for light own-account publishing, a real tax on links and reads at scale, and one more component to maintain for anyone whose actual job is not building API plumbing.