X charges $0.015 per request to create a post through its API. A post that contains a URL costs $0.200, which is 13.3 times more (X Developer Platform, retrieved 11 September 2026). Publishing links is most of what a social media scheduler exists to do.
That price appears nowhere on Buffer's pricing page, which meters channels rather than API requests (Buffer). Self-host and the registration is yours. Postiz's documentation is explicit that you create it and hold the keys (Postiz docs). So are the terms, the rate limits and the bill.
Neither side of the search results covers this. We pulled the top ten organic results for buffer alternative on 11 September 2026: eight were roundups published by a vendor that ranked its own tool in the list. Not one of the ten discussed registering a developer app. The self-hosting side has the opposite blind spot and treats the install as the whole job. The scheduler is the cheap half. The platform API access is the half you cannot buy your way out of. Self-hosting does not remove it. It hands it straight to you.
This is a spoke in Lane 01, automating your daily ops. It links up to our Ops Automation Playbook, the pillar ranking which workflows are worth attacking first.
By Dan Colta. We are a founder-led EU automation studio building owned automations for SME teams. We run our own instance of one of the tools compared here. This post comes out of operating it rather than from a vendor template.
Key Takeaways
- X charges $0.015 to create a post and $0.200 to create a post containing a URL, 13.3x for the link posts a scheduler mostly publishes (X Developer Platform). The fixed monthly subscription tiers most articles still quote are gone, replaced by pay-per-usage credits.
- Three of the four platforms gate you behind a human rather than a price. LinkedIn wants a Technical Sign Off demo (LinkedIn), Meta wants Business Verification (Meta), YouTube wants a compliance audit before any quota increase (Google). Only X will simply take your money.
- Buffer is cheap and it is the honest answer at low channel counts. Essentials is $5 per channel per month with a free tier of 3 channels (Buffer). Six channels at $360 a year beats a $299 Mixpost Pro licence plus a server in year one.
- The self-hosting case is against per-seat pricing and it starts in year two. Hootsuite's plans begin at $99 per user per month (Hootsuite) and Sprout Social Standard is $199 per seat (Sprout Social), against a Hetzner CX23 at EUR 5.49 per month (Hetzner).
- Licence, not stars, is the Postiz versus Mixpost decision. Postiz is AGPL-3.0 with 35,704 stars (GitHub); Mixpost Lite is MIT with 3,683 stars (GitHub) and lists 12 platforms (Mixpost).
Why does a self-hosted scheduler not mean fewer platform limits?
Because the limits live in the platform APIs, not in the scheduler. Postiz's own documentation says only ten of its networks need no setup. For every other one you register an app on that platform's developer portal and set the keys as environment variables on your instance (Postiz docs, retrieved 11 September 2026). Each registration is a contract you sign personally.
A hosted tool sells you a seat or a channel, never a developer app. Buffer bills per channel (Buffer), Hootsuite and Sprout Social bill per user (Hootsuite, Sprout Social). None of those pages carries a line item for API registration, platform review or per-request cost. Whoever is doing that work, you are not the party named on the developer agreement.
Self-hosting moves that to you: your app, your acceptance of the developer terms, your rate limit and your account when a platform decides to suspend something. What you get back is the post history in a database you control plus a bill that stops scaling with people.
That is the same trade every build-versus-buy decision makes, the framework in our SaaS replacement playbook. What makes schedulers unusual is that the operating work is not the server. It is the four developer relationships behind it.
What does Buffer actually cost and when is it the right answer?
Buffer Essentials is $5 per channel per month, or $60 per channel billed yearly. The free plan covers 3 channels with 10 scheduled posts each and 1 user (Buffer, retrieved 11 September 2026). Six channels is $360 a year, which is less than most people expect to read in an article about alternatives.
If you post to three channels or fewer, use the free tier and stop here. No build beats $0. Self-hosting would cost a server plus four developer app registrations to replace something you are not paying for. Buffer's price also scales with channels instead of people, which suits a small team.
Per-seat tools are where the arithmetic turns. Hootsuite's plans start at $99 for Standard and $199 for Professional, running to $399 for Advanced, per user per month on annual billing (Hootsuite, retrieved 11 September 2026). Sprout Social runs four tiers on annual billing: Essentials at $79, Standard at $199, Professional at $299 and Advanced at $399, all per seat per month (Sprout Social). Later charges flat instead of per seat, with Starter at $18.75, Growth at $37.50 and Scale at $82.50 a month on annual billing (Later).
One Sprout Social Standard seat is $2,388 a year. Three seats is $7,164. That is what a self-hosted scheduler competes against, not Buffer's $360. Worth naming what those pages put on the higher tiers, though. Hootsuite and Sprout Social both sell the upper plans on approval workflows, analytics and client reporting rather than on posting (Hootsuite, Sprout Social). The posting part is a thin layer over four APIs, the category pattern in the AI wrapper SaaS trap.
Which platforms let you register an app and which make you ask a human?
Three of the four gate you behind a person rather than a payment. LinkedIn requires a Technical Sign Off demo before production access (LinkedIn). Meta requires Business Verification for Advanced Access (Meta). YouTube requires a compliance audit before any quota increase (Google). Only X will simply take your money.
| Platform | Self-serve app creation? | Human review required? | Hard cap you cannot buy past | Marginal cost per post |
|---|---|---|---|---|
| X | Yes | No | 10,000 post creations per app per 24h, 100 per user per 15 min | $0.015 plain, $0.200 with a URL |
| Yes, Development tier | Yes, Technical Sign Off demo | 500 API calls per app per 24h, 100 per member; integration must be completed within 12 months of provisioning | No published per-call charge | |
| Yes | Yes, Business Verification for Advanced Access | 100 API-published posts per account per rolling 24h | No published per-call charge | |
| YouTube | Yes, via a Google Cloud project | Yes, compliance audit for extra quota | 10,000 units per day plus 100 uploads, by default | No published per-call charge |
Sources, all retrieved 11 September 2026: X pricing and X rate limits; LinkedIn; Meta permissions and Instagram publishing; YouTube quota.
Those caps are not comparable units. X counts requests, LinkedIn counts API calls, Instagram counts published posts and YouTube counts quota units that differ per endpoint. Read each row against itself.
X is the outlier, not in the direction its reputation suggests. Access is pay-per-usage credits with no contracts, subscriptions or minimum spend (X Developer Platform). The fixed monthly subscription tiers that page-one comparison articles still quote are no longer what X's own documentation describes. The ceiling is now the rate limit: 10,000 post creations per app per 24 hours (X Developer Platform).
LinkedIn is the row that stops most self-hosters. The default Development tier caps an app at 500 API calls per 24 hours and 100 per member. It is also time-boxed: developers are expected to build their core business use cases within twelve months of the provisioning. LinkedIn's own wording is that you must complete integration and testing inside that window (LinkedIn, documentation updated 17 August 2026). Getting past it means a scheduled demo of your product capabilities against LinkedIn's Community Management requirements, with no published turnaround for the review. Plan around that rather than resent it. For the account-level side, see LinkedIn automation in 2026.
Meta and YouTube gate on verification instead of volume. Business Verification is required for all apps requesting Advanced Access to permissions (Meta), while the publishing cap is generous at 100 API-published posts per account per rolling 24 hours (Meta). YouTube allocates 10,000 units a day plus 100 uploads by default (Google). Raising that requires an audit showing your project complies with the YouTube API Services Terms of Service (Google). An audit is not a price list. No amount of money skips it.
What does it cost to run your own scheduler for a year?
About EUR 66 for the server, plus whichever licence you choose and whatever you spend on X. A Hetzner CX23 with 2 vCPU, 4 GB RAM and 40 GB of NVMe lists at EUR 5.49 per month excluding VAT and IPv4 (Hetzner, retrieved 11 September 2026), which is EUR 65.88 over twelve months.
| Option | Year-one total | Currency | Basis |
|---|---|---|---|
| Buffer free plan | 0 | n/a | 3 channels, 10 scheduled posts per channel, 1 user |
| Postiz self-hosted on a Hetzner CX23 | 65.88 | EUR, excl. VAT and IPv4 | AGPL-3.0 licence at no cost, server only |
| Mixpost Lite on the same server | 65.88 | EUR, excl. VAT and IPv4 | MIT licence at no cost, server only |
| Mixpost Pro on the same server | 299.00 plus 65.88 | USD one-time plus EUR per year | One-time perpetual fallback licence, plus server |
| Buffer Essentials, 6 channels | 360.00 | USD | $5 per channel per month |
| Later Growth | 450.00 | USD | $37.50 per month flat on annual billing, not per channel |
| Hootsuite Standard, 1 seat | 1,188.00 | USD | $99 per user per month, annual billing |
| Sprout Social Standard, 1 seat | 2,388.00 | USD | $199 per seat per month, annual billing |
Two currencies, no conversion. Hetzner bills in euros and every other row is in dollars. Converting would bake in a rate that moves. Compare the rows you would actually pay.
Buffer wins year one at six channels. A post about alternatives should say so. $360 against a $299 Mixpost Pro licence plus EUR 65.88 of server is close enough that the exchange rate decides it. Buffer's price also covers the connection work you would otherwise do yourself. The gap opens in year two, when the only recurring line is the server.
X is the one variable you model yourself. At $0.200 per post containing a URL, one link post a day is about $73 a year and five a day about $365 (X Developer Platform). That arithmetic is ours on X's published unit price. Plain posts are $0.015, which puts the same five a day at about $27. The difference is entirely whether your posts carry links, which for most scheduling they do.
The approval step is not on this table and does not have to live in the scheduler. The approve-in-a-Slack-channel pattern adds it for the price of a webhook.
Postiz or Mixpost?
Pick on licence, not on stars. Postiz is AGPL-3.0 with 35,704 GitHub stars (GitHub, retrieved 11 September 2026). Mixpost Lite is MIT with 3,683 stars (GitHub) and Mixpost advertises 12 platforms (Mixpost). Stars measure attention. The licence decides what you may do next.
| Dimension | Postiz | Mixpost |
|---|---|---|
| Licence | AGPL-3.0 | MIT (Lite) |
| GitHub stars | 35,704 | 3,683 |
| Platforms advertised | Not stated as a single figure in the repo | 12 |
| Price basis | $0 licence, infrastructure only | Free Lite; $299 one-time Pro; $1,199 one-time Enterprise |
The AGPL clause that matters is Section 13, the network-use provision (GNU). If you modify Postiz and expose the modified version to other people over a network, you owe them the source of your changes. Running it unmodified for your own posting triggers nothing. Read it as a constraint on reselling a fork rather than on operating an instance. Take actual advice before you build a product on either licence.
MIT carries no equivalent clause, which makes Mixpost Lite the safer base if you plan to build on top of it and charge for the result. Mixpost's paid tiers are one-time payments with a perpetual fallback licence instead of subscriptions (Mixpost), unusual in this category and a genuine point in its favour.
Neither project publishes hardware requirements. As of 11 September 2026 neither repository's README nor its documentation states a RAM or CPU minimum. Treat any figure you see quoted as somebody's install rather than a spec. If you were going to wire the platform APIs yourself anyway, where a no-code tool sits between building and buying is the comparison to read first.
What actually broke when we ran one
Three things, none of them the server. We run Postiz at social.nodesparks.com on a Hetzner CX23, 2 vCPU and 4 GB RAM at EUR 5.49 per month in Falkenstein (Hetzner), live since 27 August 2026. LinkedIn, X and YouTube are connected. Reddit is not.
The failure mode is credentials, not infrastructure. X refused to connect behind a generic "Could not connect to the platform" message while the backend returned HTTP 200 and the container logs stayed clean. The cause was a dead Consumer Key pair: the keys had been regenerated in X's developer portal after they were copied into the environment file. Every visible setting looked correct and none of it worked. A single client_credentials call against X's own token endpoint isolates that in seconds. Owning the app registration means owning the debugging, with error messages worse than what a vendor who has already fielded that failure would show you.
YouTube needed two Google APIs enabled, not one. Both the Data API and the Analytics API have to be on in the Google Cloud project. The tool surfaced the failure as "we couldn't find any YouTube channels", which points at the wrong thing entirely. The container logs were the only honest signal.
Reddit is the channel we could not connect. We could not complete the app registration. We also could not verify Reddit's current API terms against a primary source at the time of writing, because Reddit's own documentation and support domains block automated fetching. The honest statement is the narrow one: it did not work for us. We are not asserting why. If you need Reddit, check its terms yourself before planning around it.
None of this is an argument against self-hosting. The server is boring. The four OAuth relationships are not.
When should you self-host and when should you not?
Self-host when you are paying per seat or per channel at volume you control. One Sprout Social Standard seat is $199 per month, or $2,388 a year (Sprout Social), against EUR 65.88 for a server that does not count logins. Three channels on Buffer's free plan is the opposite case entirely (Buffer).
Four signals to build:
- The seats are the cost. Hootsuite Standard at $99 per user per month makes everyone who touches the calendar a line item. A self-hosted instance does not count logins.
- You post to more than six channels. Buffer's per-channel price is fair and still linear. Twelve channels on Essentials is $720 a year against one flat server bill.
- You already run a server and somebody patches it. One more container costs close to nothing. If this would be your first server, it does not.
- You want the post data in your own database. Scheduled content and per-channel results in Postgres you control are not something any of the hosted tiers compared here offers.
Four signals to keep renting:
- Three channels or fewer. Buffer's free plan covers 3 channels, 10 scheduled posts each and 1 user. Nothing you build beats that.
- Nobody owns patching, backups and token renewal. Tokens expire and platforms change scopes. A scheduler that quietly stops publishing is worse than one you pay for, because you find out from an empty calendar three weeks later rather than from an alert.
- You need LinkedIn production access on a deadline. The Technical Sign Off demo has no published turnaround. Treat that wait as unbounded.
- What you want is approvals, analytics or client reporting. Those are the features the per-seat tiers genuinely sell. The open-source projects are thinner there.
The hybrid is usually the right shape: Buffer's free tier for the three channels that matter, a self-hosted instance for the volume.
FAQ
Six questions are answered in the schema above: what one is, whether it beats Buffer on cost, what X charges per post, whether you need your own developer app per platform, how the Postiz and Mixpost licences differ and who should not self-host.
The bottom line
The scheduler is the cheap half. Postiz costs nothing to licence, Mixpost Lite costs nothing to licence and a server that runs either is EUR 5.49 a month. If that were the whole decision, it would not be a decision.
The platform API access is the other half and it has no price tag. Three of the four platforms will not sell you access at any number. LinkedIn wants a demo. Meta wants Business Verification. YouTube wants a compliance audit. X will take your money, at $0.015 a post and $0.200 if that post carries a link. Self-hosting hands you all of it under your own developer account, with worse error messages.
So the rule is small. Three channels means Buffer's free tier. If you pay per seat, or per channel at real volume, or you want the post history in a database you own, self-host and budget for four developer relationships rather than one server.
If you want a second read on which of those developer app registrations you would actually have to own, or a scoped estimate for building the pieces the open-source projects leave out, our team can help. Start with the Ops Automation Playbook for the full ranking of which ops workflows to automate first.
Sources (retrieved 2026-09-11):
- X Developer Platform, X API pay-per-usage pricing and credits: https://docs.x.com/x-api/getting-started/pricing
- X Developer Platform, Rate limits: https://docs.x.com/x-api/fundamentals/rate-limits
- LinkedIn (Microsoft Learn), Increasing Access: https://learn.microsoft.com/en-us/linkedin/marketing/increasing-access
- LinkedIn (Microsoft Learn), Integration Requirements for Community Management: https://learn.microsoft.com/en-us/linkedin/marketing/community-management/integration-requirements-community-management
- Meta for Developers, Permissions Reference: https://developers.facebook.com/docs/permissions
- Meta for Developers, Content Publishing, Instagram Platform: https://developers.facebook.com/docs/instagram-platform/content-publishing
- Google, YouTube Data API Overview, Getting Started: https://developers.google.com/youtube/v3/getting-started
- Google, Quota and Compliance Audits, YouTube Data API: https://developers.google.com/youtube/v3/guides/quota_and_compliance_audits
- Postiz Documentation, Provider API keys: https://docs.postiz.com/providers
- GitHub, gitroomhq/postiz-app: https://github.com/gitroomhq/postiz-app
- GitHub, inovector/mixpost: https://github.com/inovector/mixpost
- Mixpost pricing: https://mixpost.app/pricing
- Mixpost, supported platforms: https://mixpost.app/
- Buffer pricing: https://buffer.com/pricing
- Hootsuite pricing: https://www.hootsuite.com/pricing
- Sprout Social pricing: https://sproutsocial.com/pricing/
- Later pricing: https://www.later.com/pricing/
- Hetzner, Price Adjustment 15 June 2026: https://docs.hetzner.com/general/infrastructure-and-availability/price-adjustment/
- GNU, GNU Affero General Public License v3.0: https://www.gnu.org/licenses/agpl-3.0.html
Frequently asked questions
What is a self-hosted social media scheduler?
A self-hosted social media scheduler is a scheduling app you run on your own server instead of renting as SaaS. Postiz and Mixpost are the two main open-source options, licensed AGPL-3.0 and MIT respectively. You supply the infrastructure. A Hetzner CX23 with 2 vCPU and 4 GB RAM lists at EUR 5.49 per month excluding VAT and IPv4. You also supply the platform access: Postiz's own documentation states that most platforms need you to register a developer app first and put the keys in your environment.
Is a self-hosted buffer alternative actually cheaper than Buffer?
Often not, at low channel counts. Buffer Essentials is $5 per channel per month with a free tier covering 3 channels. Six channels costs $360 a year. Mixpost Pro is a $299 one-time licence plus a server at EUR 5.49 per month, which lands in the same range in year one and only pulls ahead from year two. Self-hosting wins clearly against per-seat pricing: Hootsuite's paid plans start at $99 per user per month and Sprout Social Standard is $199 per seat per month.
What does it cost to post to X through the API?
X charges $0.015 per request to create a post and $0.200 per request to create a post that contains a URL, 13.3 times more for exactly the kind of post a scheduler publishes. X's developer pricing page states there are no contracts, subscriptions or minimum spend. Access is pay-per-usage credits rather than the fixed monthly subscription tiers that most comparison articles still quote. On X's published unit price, one link post a day works out at roughly $73 a year.
Do you need your own developer app for each platform?
For most of them, yes. Postiz's documentation tells self-hosters to register an app on each platform's developer portal and set the keys as environment variables, which means you personally accept each platform's developer terms and hold its rate limits. LinkedIn's default Development tier allows 500 API calls per app per 24 hours, and LinkedIn expects you to complete your integration and testing within twelve months of receiving that access. Instagram caps accounts at 100 API-published posts per rolling 24 hours. YouTube's Data API allocates 10,000 units plus 100 uploads per day.
Postiz or Mixpost: which licence should you pick?
The licence is the real difference between them. Postiz is AGPL-3.0 with 35,704 GitHub stars. Its Section 13 network clause applies if you modify it and then expose the modified version to other people. Running it unmodified for your own posting does not trigger that. Mixpost Lite is MIT with 3,683 stars, which is the more permissive base if you plan to fork it into something you sell. Mixpost Pro is a $299 one-time licence and Enterprise is $1,199.
Who should not self-host a social media scheduler?
Anyone posting to three channels or fewer, because Buffer's free plan already covers 3 channels, 10 scheduled posts per channel and 1 user at no cost. Anyone with nobody to own patching, backups and OAuth token renewal, because a scheduler that silently stops publishing is worse than one you pay for. Anyone who needs LinkedIn production API access on a deadline, since LinkedIn requires a Technical Sign Off demo and publishes no turnaround time. Anyone whose real need is approvals and reporting rather than posting.