What the X (Twitter) API Actually Costs in 2026: Pay-Per-Use Pricing Explained
Key Takeaway: X killed its subscription tiers in early 2026 and switched to pay-per-use pricing. There are no more $200/month or $5,000/month plans for new users. You buy credits upfront and pay per resource: $0.005 per post read, $0.01 per user profile, $0.01 per post created. There is a hard cap of 2 million post reads per month unless you negotiate an Enterprise deal. For most read-heavy workloads, third-party providers now cost a fraction of the official API.
Last updated: March 23, 2026
Table of Contents
- How X API Pricing Works Now (Pay-Per-Use)
- What Happened to Basic and Pro Plans?
- How Much Does the X API Really Cost? Real-World Scenarios
- When the Official X API Is the Right Choice
- When Alternatives Make More Sense
- How Third-Party X API Providers Compare on Price
- The Flat-Rate vs. Per-Resource Pricing Difference
- How to Reduce Your X API Costs
- FAQ
--
How X API Pricing Works Now (Pay-Per-Use)
As of early 2026, X has moved to a consumption-based model. No subscriptions, no monthly commitments. You purchase credits through the Developer Console, and those credits are deducted in real time as you make API calls.
Different endpoints have different per-unit costs. Here is the full breakdown straight from X's official pricing page:
| Resource | Unit Cost | Charged Per |
|---|---|---|
| Posts: Read | $0.005 | resource fetched |
| Users: Read | $0.010 | resource fetched |
| Following/Followers: Read | $0.010 | resource fetched |
| DM Events: Read | $0.010 | resource fetched |
| Content: Create (posting) | $0.010 | request |
| DM Interaction: Create | $0.015 | request |
| User Interaction: Create (like, follow) | $0.015 | request |
| Lists: Read | $0.005 | resource fetched |
| Spaces: Read | $0.005 | resource fetched |
| Communities: Read | $0.005 | resource fetched |
| Analytics: Read | $0.005 | resource fetched |
| Trends: Read | $0.010 | resource fetched |
| Media: Read | $0.005 | resource fetched |
A few things to notice. Read operations for core resources (posts, lists, spaces) cluster around $0.005. User-related reads and write operations cost $0.010-$0.015. There is no flat per-request fee; you pay for each individual resource returned. If a search query returns 20 posts, that is 20 x $0.005 = $0.10.
X also provides auto-recharge (set a threshold, credits top up automatically), spending limits per billing cycle, and a real-time Usage API endpoint so you can monitor consumption programmatically.
The 24-Hour Deduplication Rule
If you fetch the same resource (a specific post, a specific user profile) more than once within a UTC day, you are only charged for the first request. The deduplication window resets at midnight UTC. X calls this a "soft guarantee" since edge cases like service outages may prevent deduplication. In practice, this means repeat reads of the same data are free for the rest of the day.
The 2 Million Post Reads Cap
This is the detail most pricing guides skip. Pay-per-use accounts are capped at 2 million post reads per month. That sounds like a lot until you do the math for a real monitoring workflow (more on that below). If you need more than 2M reads, your only option is Enterprise pricing, which means contacting X's sales team, negotiating custom terms, and paying significantly more. Approval is not guaranteed and the process can take weeks.
Free xAI API Credits: Is the Kickback Worth It?
X now bundles xAI (Grok) API credits with X API spending. The tiers:
| Cumulative X API Spend (per billing cycle) | xAI Credit Rate |
|---|---|
| $0 - $199 | 0% |
| $200 - $499 | 10% |
| $500 - $999 | 15% |
| $1,000+ | 20% |
At $1,000 in X API spend, you receive $200 in xAI credits. The rate applies retroactively to your full balance, but you only receive the incremental difference as you cross each threshold.
Honest assessment: this is a nice perk if you are already building with Grok or xAI models. It is not a reason to spend more on the X API than your workload requires. A 20% kickback on a $1,000 bill still means you spent $800 net on X API credits.
--
What Happened to Basic and Pro Plans?
If you are reading older guides that reference a $200/month Basic plan or a $5,000/month Pro plan, that information is outdated.
Here is what actually happened:
| Date | Change | Impact |
|---|---|---|
| Late 2022 | Elon Musk acquires Twitter | Speculation about API monetization begins |
| Feb 2023 | Free API access eliminated, paid tiers introduced | Third-party clients shut down overnight |
| Mar 2023 | Basic ($100/mo), Pro ($2,500/mo), Enterprise ($42K+/mo) launched | Developer ecosystem fragments |
| Mid-2024 | Basic doubles to $200/mo | Indie developers priced out |
| Late 2025 | Pay-per-use beta announced | $500 vouchers distributed to testers |
| Early 2026 | Pay-per-use goes live as the default model | Subscription tiers become "legacy" |
New users can only access the pay-per-use model. If you sign up for the X API today, there is no option to buy a Basic or Pro subscription. Existing subscribers on legacy plans can stay on them or opt into pay-per-use through the Developer Console. They can also switch back to their legacy plan if they change their mind.
Having tracked X API pricing changes since the v1.1 era, this is the most significant restructuring since the 2023 paywall went up. The old model punished anyone who fell between tiers. The new model charges for what you use, but the per-resource costs add up fast at scale, and the 2M post cap introduces a hard ceiling that did not exist on the old Pro plan.
The broader context matters too. X is not the only platform monetizing data access aggressively. Reddit struck a $60 million annual deal with Google for AI training data. Social media data has become a revenue center, not an afterthought, and pricing reflects that across the industry.
--
How Much Does the X API Really Cost? Real-World Scenarios
Per-endpoint pricing tables are useful, but what actually matters is: how much will I spend on a real workflow? Let's run the numbers.
Brand Monitoring: 10K Posts + Author Profiles Per Month
You want to track mentions of your brand and analyze who is talking about you.
- 10,000 post reads: 10,000 x $0.005 = $50
- 10,000 user reads (one per post author): 10,000 x $0.010 = $100
- Total: $150/month
That is cheaper than the old $200 Basic plan, and you get more flexibility. But note: on the old Basic plan, user reads were bundled. On pay-per-use, every resource type is billed separately.
Competitor Analysis: 5 Competitors, Weekly
You pull each competitor's recent posts and mentions weekly.
- 5 competitors x 200 posts/week x 4 weeks = 4,000 post reads: $20
- 4,000 user reads (post authors): $40
- 5 competitor profile lookups x 4 weeks = 20 user reads: $0.20
- Total: $60/month
Affordable. This is one scenario where pay-per-use genuinely works well for small-to-medium workloads.
Large-Scale Research: 500K Posts
An academic study, ML training dataset, or market research project requiring 500,000 posts.
- 500,000 post reads: 500,000 x $0.005 = $2,500
- If you also need author profiles: add 500,000 x $0.010 = $5,000
- Total: $2,500 - $7,500 depending on scope
And you are using 25% of your monthly 2M post cap on a single project. If you have other workloads running simultaneously, that cap becomes a constraint fast.
Real-Time Monitoring: Continuous Polling
Polling a search endpoint every 30 seconds, 24/7, with each call returning 20 posts.
- 2 requests/minute x 60 x 24 x 30 = 86,400 requests/month
- 86,400 x 20 posts = 1,728,000 post reads: $8,640
- Already approaching the 2M cap
This is where pay-per-use gets expensive. The old Pro plan at $5,000/month included 1M post reads with full-archive search. Continuous monitoring at moderate frequency now costs more on pay-per-use. For alternative approaches to real-time monitoring that avoid this cost structure, third-party providers offer significantly lower per-item rates.
| Scenario | Monthly Post Reads | X API Cost (Posts Only) | Total with User Reads |
|---|---|---|---|
| Brand monitoring (10K posts) | 10,000 | $50 | $150 |
| Competitor analysis (5 targets) | 4,000 | $20 | $60 |
| Research project (500K posts) | 500,000 | $2,500 | $2,500 - $7,500 |
| Real-time monitoring (24/7 poll) | 1,728,000 | $8,640 | $8,640+ |
--
When the Official X API Is the Right Choice
The official API is the only option when you need to write to the platform. Posting tweets, sending DMs, managing lists, liking, following, deleting content. No third-party provider can do this for you.
Other scenarios where the official API makes sense:
- Ads API integration. Campaign management and analytics require official API access.
- Compliance requirements. Regulated industries (finance, healthcare) may need a direct data agreement with X.
- Real-time filtered streams. X's streaming endpoints push data to you as it happens. If sub-second latency is critical, the official firehose is hard to replace.
- Low-volume read workloads. If you need fewer than 5,000-10,000 post reads per month, pay-per-use is genuinely cheap ($25-$50).
--
When Alternatives Make More Sense
Third-party X data providers use their own infrastructure to serve Twitter/X data through independent REST APIs. They charge differently, usually per request or per batch of results, and they are not bound by X's per-resource pricing model.
Alternatives typically win in three situations.
Read-heavy workloads at scale. If your application primarily reads posts, profiles, and followers without needing to post or interact, you are paying the X API's per-resource premium for write capabilities you never use.
Budget constraints on medium volume. The awkward zone between "this is cheap" and "this is Enterprise" on the official API (roughly 50K-2M post reads/month) is exactly where third-party providers offer the biggest savings.
Simplicity. The official X API requires OAuth 2.0 authentication, credit management, spending limit configuration, and monitoring multiple per-endpoint rates. Most alternatives authenticate with a single API key header and use a flat per-request pricing model.
--
How Third-Party X API Providers Compare on Price
Disclosure: Sorsa API is our product. We have aimed to keep this comparison balanced, but recommend testing any solution with your own workload before committing.
The fundamental difference between the official X API and most third-party providers is the billing unit. X charges per resource fetched (each post, each user profile, each follower in a response is a separate billable unit). Providers like Sorsa API charge per request (one API call = one unit from your quota, regardless of how many items it returns).
This distinction has a massive impact on cost.
| Official X API (Pay-Per-Use) | Sorsa API | Typical Third-Party Provider | |
|---|---|---|---|
| Pricing model | Per resource fetched | Per request (flat) | Varies (credits, per-tweet, per-request) |
| Cost: 1K post reads | $5.00 | $0.10 (50 requests x $0.00199) | $0.15 - $0.50 |
| Cost: 1K user profiles | $10.00 | $0.01 (5 requests via /followers) | $0.18 - $1.00 |
| Monthly cap | 2M post reads | Based on plan (10K-500K requests) | Varies |
| Rate limit | 300 req/15 min (varies by endpoint) | 20 req/s (all plans) | Varies |
| Auth method | OAuth 2.0 + Bearer Token | Single API key header | Usually API key |
| Write access | Yes (post, DM, like, follow) | No (read-only) | Rarely |
| Batch endpoints | Limited | Yes (100 tweets or profiles per call) | Sometimes |
Sorsa Pro plan ($199/mo for 100K requests). Search and timeline endpoints return 20 results per request. Follower endpoints return up to 200 profiles per request.
The numbers tell the story. Collecting 100,000 tweets on the official API costs $500 in post reads alone. On Sorsa's Pro plan, 5,000 requests to /search-tweets (20 tweets each) cover the same volume for about $10. And each of those tweet responses includes full author profiles at no extra charge, so you get user data for free that would cost another $1,000 on the official API.
When I helped a fintech client migrate their social sentiment pipeline off the official Pro plan last year, the cost dropped from around $5,000/month to under $200. They kept a minimal official API setup for posting alerts to their company X account but moved all read operations (mention tracking, competitor monitoring, follower analysis) to a third-party provider. That hybrid approach cut their total API spend by over 90%. The client asked to remain anonymous due to NDA terms, but the pattern is one I have seen across dozens of similar migrations. For step-by-step instructions, see the migration guide.
For a full walkthrough of getting started, see the Sorsa API quickstart guide. For a detailed comparison of all major third-party providers, see our X API alternatives guide.
--
The Flat-Rate vs. Per-Resource Pricing Difference
This pricing distinction deserves its own section because it is the single biggest factor in cost differences between providers.
X API (per-resource): You call a search endpoint. It returns 20 posts. You are charged for 20 post reads ($0.10). If you also expand user data, you are charged for 20 user reads ($0.20). Total cost for one search call: $0.30.
Flat-rate provider (per-request): You call a search endpoint. It returns 20 posts with full author profiles embedded. You are charged for 1 request. On Sorsa's Pro plan, that is $0.00199. The 20 tweets and 20 user profiles cost the same as a single ID conversion.
Now scale that to batch endpoints. Sorsa's [/tweet-info-bulk](https://docs.sorsa.io/api-reference/tweets/tweet-data-batch) accepts up to 100 tweet IDs in a single request. One request, $0.00199, 100 complete tweets with author data. The same operation on the official API: 100 post reads ($0.50) + 100 user reads ($1.00) = $1.50.
That is a 750x cost difference on the same data.
This does not mean flat-rate is always better. If you only need 100 post reads per month, the official API costs $0.50 total and there is no reason to pay $49/month for a third-party plan. The economics flip once your volume crosses roughly 10,000 post reads per month.
--
How to Reduce Your X API Costs
These strategies apply regardless of which provider you use.
Cache aggressively. On the official API, the 24-hour deduplication window helps, but caching on your end eliminates redundant calls entirely. Store post data with a 6-12 hour TTL. User profiles change rarely; a 24-hour cache is reasonable.
Batch whenever possible. If an endpoint accepts multiple IDs, use it. Fetching 100 tweets individually costs 100x more than one batch call on any provider. On the official API, batching reduces the number of billable requests even though per-resource charges still apply.
Use field selection on the official API. The tweet.fields and user.fields parameters let you request only the data you need. Smaller responses mean faster processing and less wasted bandwidth, though the per-resource charge stays the same.
Poll less, stream more. If you are checking for new posts every 60 seconds, switch to X's filtered stream endpoint (official API) or use list-based monitoring on third-party providers. Polling 50 accounts individually costs 50x more than monitoring them through a single X List. For Python implementation examples of both approaches, see our Twitter API Python guide.
Right-size your plan. On the official API, this means keeping spending limits tight during development. On third-party providers, compare cost-per-request across plan tiers. Moving from Sorsa Starter ($49/mo) to Pro ($199/mo) cuts the per-request cost from $0.0049 to $0.00199, a 59% reduction. If you consistently use 8,000+ requests on Starter, upgrading saves money. See the optimization guide for more patterns.
--
FAQ
Is there a free X (Twitter) API in 2026?
Not in any meaningful sense. The old free tier (which gave limited read access) has been replaced by the pay-per-use model. There is no subscription fee to start, but you must purchase credits before making any API call. There is no free credit allowance. If your credit balance hits zero (or goes slightly negative), API requests are blocked until you add funds.
For zero-cost access to X data, your options are third-party providers that offer free tiers or trials, or manual collection through X's web interface. Scraping is another route, though it comes with its own costs and risks; see our guide to scraping X and managed scrapers comparison. Programmatic free access to the official API no longer exists.
Why did the Twitter API become so expensive?
Three converging forces. First, Musk's acquisition was heavily debt-financed, and X needed new revenue streams. API access was an obvious monetization target. Second, AI companies were consuming massive volumes of social data for model training. X, like Reddit and others, recognized the commercial value of its data corpus. Third, free API access had enabled spam bots, scrapers, and tools that X wanted to curtail. Aggressive pricing served as both a revenue strategy and an access control mechanism.
Can I still get the $200 Basic or $5,000 Pro plan?
No, if you are a new user. The Basic and Pro subscription tiers are legacy plans available only to developers who subscribed before the pay-per-use transition. Legacy subscribers can stay on their current plan or switch to pay-per-use through the Developer Console. If they switch, they can also revert back. But you cannot sign up for a new Basic or Pro subscription in 2026.
What happens if I hit the 2 million post reads cap?
Your API requests for post-read endpoints are blocked until the next billing cycle. Other endpoint types (user reads, content creation) are not affected by the post-read cap, but they have their own usage limits. For a full breakdown of per-endpoint rate limits (separate from billing caps), see our X API rate limits guide. The only way to exceed 2M post reads per month is to contact X's enterprise sales team and negotiate a custom plan. Approval is selective, the process takes time, and pricing is not public, though historically Enterprise has started around $42,000/month.
--
Daniel Kolbassen is a data engineer and API infrastructure consultant. He has worked with the Twitter/X API since the v1.1 era and has helped over 40 companies restructure their data pipelines after the 2023 pricing overhaul. Follow him on Twitter/X or connect on LinkedIn.