Posting API or finished product: what you are actually buying

What you actually buy when you buy an API

Every social network has its own API, its own formats, its own OAuth and its own habit of changing the rules without much warning. A unified posting API solves that specific problem: one call, one data schema, one place where things break. It is a real and large saving, and anybody who has integrated Instagram and TikTok by hand knows exactly what it is worth.

What you are not buying is a product. The API returns a post identifier, not an interface where a community manager drags a piece onto a Tuesday. Written down like that the distinction looks obvious, and it is forgotten with great consistency at budgeting time, because the subscription cost is printed and the cost of what is missing is not.

Two things also get confused in this list and should not be. Five of the six providers publish to social networks. Unipile does not: it is a messaging and communications API covering LinkedIn, Instagram, Telegram, WhatsApp, email and calendars, and its documentation mentions no post scheduling at all. It appears here because it competes for the same budget in the same conversations, not because it does the same job. If what you need is public posts, that entry is not yours however attractive its per account price looks; if what you need is reading and writing conversations, it is probably the only one of the six that fits.

Six providers, six ways of charging

Prices checked on 8 September 2026: Ayrshare, Blotato, upload-post, bundle.social, Unipile, Post Bridge.

The figures are in the currency each provider publishes them in and none have been converted. Unipile publishes both euros and dollars; the rest publish dollars only. Read the entry tier column with some suspicion, because two of those zeroes are free plans with real caps rather than a way to run anything.

The cost by how many accounts you connect

This is the table that decides things, because the entry price misleads. What matters is the slope: what the number of accounts you will have in two years costs, not the three you start with.

Three readings of that table. The first: Ayrshare has no figure because its unit is the social profile, and a single profile holds up to fourteen accounts on different networks, so "account" and "profile" are not the same thing and its page publishes no conversion. That is not opacity, it is a different unit.

The second: bundle.social and Post Bridge do not scale per account at all. On the first the real limit is monthly posting volume, 10,000 on its middle plan and 100,000 on the top one, and one network sits outside the flat rate and is billed by usage. On the second the top tier carries unlimited accounts, so a hundred accounts and a thousand cost the same. If your product connects many accounts each doing little volume, those two models are substantially cheaper than any per account price.

The third: Unipile is the only one publishing a banded ladder up to 5,000 accounts and beyond, with the per account price falling as you grow. It is the most predictable of the six at scale, and also the only one on this list that does not publish posts.

What is still yours afterwards

Here is the real budget, and it comes out of the providers' own documentation. None of the six hands you these pieces, and in a real product every one of them has a user in front of it.

• The calendar and the editor. The API accepts a scheduled date; it does not draw a month, drag pieces around, show conflicts or warn you that Tuesday has three things at the same hour.
• The approvals. Not one of the six documents a review or approval flow. If your product has a client who says yes before anything goes out, that state, that record and that screen are yours to write.
• The analytics history. Several expose per post metrics, but none is a warehouse. The twelve month series your user wants to see is stored by you, with its aggregation rules, which are not obvious.
• The media library. Blotato requires the file to be uploaded to its own endpoints before publishing, upload-post processes video but documents no library with an interface, and in every case managing reusable assets is on you.
• The inbox. Some expose comments and messages over the API, with uneven coverage: on Blotato, for instance, on two networks only. An inbox is another thing entirely: threads, states, assignment and a clock.
• The retries. Blotato's documentation says explicitly that a failed post needs manual intervention. That logic, with its queues and its alerts, is your code from day one.

An honest way to size it is to count screens. Calendar, editor, library, inbox, analytics dashboard, connected account management and user settings are seven surfaces, and none of them is a form. Before deciding, it is worth reading the social media stack cost guide, which does the same sum from the other side.

The limits you do not control

There is one more layer that appears on no price list and that decides how reliable whatever you build on top will be.

Rate limits come in two kinds. The provider's own are published and manageable: Ayrshare allows 300 requests every five minutes per profile and suspends a profile that racks up a thousand limit errors in a day; bundle.social documents bursts of 100 requests a second and 2,000 a minute; Post Bridge caps 100 posts an hour per user. The ones that hurt are the others, the networks' own: 50 posts a day on Instagram, 25 per Page on Facebook, 10 pins a day on Pinterest. Nobody negotiates those, and your product has to explain them to a user who does not understand why their post did not go out.

Tokens are the expense that never ends. Some providers renew OAuth for you and others hand you back a disconnected account and leave you the reconnection screen. Every network changes its rules on its own schedule, so that maintenance never finishes and has to be budgeted as a standing line rather than as a project. What holding fourteen integrations up at once involves is told from the inside in publishing at scale.

Then there is the availability commitment. Of the six, only Ayrshare publishes a formal SLA, at 99.9 %, and only for its Enterprise tier. The rest publish none. Unipile states 99.9 % observed uptime over twenty four months and its own documentation clarifies that continuous provider connectivity is not guaranteed. If your contract with your client promises something, you are the one promising it.

How to decide without kidding yourself

The question is not which of the six is best. It is what you are building.

If you are making a software product and publishing is a feature inside it, buy the API. Choose on billing unit rather than entry price: if you will connect many accounts each doing little, look at the flat rates with unlimited accounts; if you will have few accounts doing a lot, look at the posting caps before the account counts; and if you are heading for thousands, look at who publishes a banded ladder. Then add the seven screens and the token maintenance, and you have the real budget.

If you run an agency and what you need is to publish for your clients, you are not in this market, however tempting these entry prices look. What you need is precisely the thing these APIs are an ingredient of, with the calendar, the approvals and the inbox already built. GoFeed is a product of that second kind rather than an API you can integrate into your own, and framing it any other way would mislead you. Which networks it covers and in which formats is in the connected networks guide.

And there is a third road plenty of people rule out too early: open source that already solves the product half and runs on your own server. It has costs of its own, and they are not zero. The honest version of that comparison is in the Postiz comparison.