Postiz Self-Hosted and the Credential Wall: The Vendor Itself Says Approval Takes More Than a Month

The docker compose comes up in about four minutes. You get a login screen, a calendar, a workspace, and a very satisfying feeling that you have just deleted a subscription from your life. Then you click "add channel," and the application asks you for a client ID and a client secret that only the platform can issue you, and the satisfying feeling ends.

This is the wall every Postiz self-hoster meets, and the honest thing about it is that you do not have to take our word for the difficulty. The vendor concedes the whole point on its own pricing page. We fetched it again today, and the sentence is still there.

"Facebook, Instagram, Threads and YouTube are more complicated and can usually take more than a month with multiple changes in the submissions, privacy policy and terms of service."

That is Postiz, describing what happens after you have finished the part everyone writes tutorials about. It is a rare thing for a vendor to publish, and it is worth more than any third-party estimate, because nobody has an incentive to overstate the pain of the free option they themselves maintain.

The deployment is solved. That was never the hard part.

Credit where it is due: Postiz has made the infrastructure side genuinely undemanding, and it says so in plain language. Asked where you can run it, today's FAQ answers: "We can deploy it on any server you want, it must be a server that you owe for example: AWS, Azure, GCP, Railway, Hetzner, Coolify, etc."

That sentence is also the reason for a search pattern that looks odd at first. People type coolify postiz, but the phrase returns an empty autocomplete in both word orders — almost nobody is typing it, and there is nothing to optimise for. Yet the pairing keeps appearing in forums, and the explanation is simply that Postiz named Coolify itself. The vendor put the word in front of people; the search volume never followed. Worth remembering the next time a keyword tool shows you an intriguing combination with no traffic behind it.

The self-host cluster that does get typed is straightforward: postiz self host branches into self hosted, free, features, docs, requirements, tiktok, alternative, api and github. Notice what that list is. Four of those are deployment questions, and the rest are the questions people ask once deployment worked and they hit the credential wall.

About TikTok specifically, and what we will not make up

The keyword that brought most people here is postiz self hosted tiktok, so here is the precise state of the evidence. In today's fetch, Postiz names four platforms as the complicated ones: Facebook, Instagram, Threads and YouTube. TikTok is not in that list. We also found no statement from Postiz anywhere on today's pass giving a timeline, a difficulty rating or a process description for TikTok credentials in particular.

There is no shortage of blog posts willing to tell you how long TikTok takes. We are not going to join them, because a number we cannot trace to the platform's own announcement or to a vendor's first-hand statement is worse than no number: it looks authoritative, it gets copied, and six months later four articles are citing each other. What we can tell you from the vendor's own text is the shape of the work rather than its duration — submissions, a privacy policy, terms of service, and multiple rounds of iteration. That shape is the same regardless of which platform is making you do it.

If you need an actual answer for TikTok, the reliable path is to open a developer application yourself and watch the clock. That is slower than reading a blog post and it is the only version that is true for your account, in your region, this month. We wrote up the one vendor who finally put a number on approval times, and that number remains the only first-hand figure we have found worth citing.

What you are actually bidding against

Self-hosting is a trade, and a trade needs both sides on the table. The thing you save is the hosted subscription, so here is what that subscription costs today, with the three columns that make a pricing table usable rather than decorative.

The column-equality check passes — $29 / $39 / $49 / $99 does not track the channel column or either AI column cell for cell, so nothing here is a misread quota. And two things fall out that are worth carrying.

First, the marginal column is not monotonic. The step from Team to Pro is the cheapest channel you will ever buy from this vendor at fifty cents each, and then the step to Ultimate gets more expensive again at seventy-one cents. Averages hide that completely: the average price per channel falls smoothly the whole way down the table while the real cost of your next channel goes down and then back up. If you are at eleven channels, the correct move is to jump straight to Pro, and it is not close.

Second, Postiz's cards carry no quantity selector — no slider, no dropdown for channel count — and yet the curve is purely decreasing. That is a useful counter-example to the usual first glance at a pricing page. A selector is a reliable signal that rates will fall; the absence of one is only a weak hint that they will not. On the same day, from the same kind of page, one vendor with no selector runs purely downhill and another runs monotonically uphill. Check the curve, do not infer it from the widget.

The three things we still could not verify, stated plainly

This matters more than anything above, because the gaps are exactly where a confident-sounding article would quietly invent something.

One: the price of app-review assistance. Postiz offers to help get your apps approved, and says explicitly that you are not obliged to pay for approval itself — "You don't have to pay for apps approval... We have provided additional services to help you with that." The FAQ then refers to "different prices between social media apps approval" depending on how complicated the platform is. It shows no figure. This is the second round running that we have gone looking for that number and not found it, which means the honest summary of the self-host economics is: you save a known subscription and take on an unknown one.

Two: feature parity between hosted and self-hosted. Three rounds now without a verified statement. We are not going to assert they are identical and we are not going to assert they differ. If a specific feature is load-bearing for you, ask before you migrate.

Three: annual billing amounts. Not rendered today, so every figure in the table above is monthly and should be treated that way.

None of this is a verdict on Postiz, which publishes more about its own friction than most vendors in the category do. Our fuller accounting of what self-hosted Postiz costs is here, and the separate question of what the subscription is really buying is here. This piece is narrower: it is about the one sentence the vendor wrote that makes the whole trade legible.

FAQ

Do I need my own API credentials to self-host Postiz?

Yes. That is the structural difference between the hosted and self-hosted versions: hosted means you are borrowing the vendor's approved platform apps, self-hosted means you register your own with each platform. The software is the same download either way; the credentials are the part you cannot copy.

How long does approval actually take?

For Facebook, Instagram, Threads and YouTube, Postiz's own FAQ says usually more than a month, with multiple iterations on submissions, privacy policy and terms of service. For TikTok, we have no first-hand figure from Postiz or from the platform, and we are not going to estimate one. Assume the same shape of work — a review queue you cannot accelerate — and plan your launch date around it rather than against it.

So is self-hosting actually cheaper?

It depends entirely on what your time is worth and whether you need the hard platforms. If you only need the platforms with light review and you enjoy running your own stack, the subscription genuinely disappears. If you need the four the vendor flags as complicated, you have swapped a monthly bill for a month of paperwork plus a privacy policy plus an unpublished assistance fee if you decide not to do it yourself. Set against a vendor whose entire proposition is that credentials are someone else's problem, that is the real comparison.

What to take away

Before you run the compose file, make a list of the platforms you actually need, and split it in two: the ones with light credential review, and the ones the vendor itself flags as month-plus. If everything you need is in the first list, self-hosting is close to free and you should do it. If anything load-bearing is in the second, you have not removed a cost — you have converted a recurring, predictable, cancellable bill into a one-off queue you cannot hurry, followed by ongoing maintenance of an app review that platforms periodically reopen. Both can be the right answer. Only one of them is the one people imagine when they type "self hosted, free."