ManyChat Pricing Changes: What Actually Moves Your Bill

This article was commissioned as a timeline of ManyChat pricing changes — a tidy chronology of every increase since launch. We could not build it honestly, so we didn't. What follows is the accounting of why, what we could actually verify, and the price change that hits most accounts without appearing on any timeline because nobody announces it.

If you came here for a table of dated increases, we're about to disappoint you and then, hopefully, be more useful than that table would have been. The short version, if you only read one line: check your own invoices, because they are the only pricing history that is definitionally true for you.

What we set out to build, and why we scrapped it

The plan was straightforward: read the published tiers, compare against archived versions, and chart the changes. The blocker showed up immediately and it was not subtle.

We could not establish what ManyChat charges *today* with enough confidence to publish a number — let alone what it charged in 2021. In mid-July 2026 we found ManyChat's own community FAQ stating an entry Pro price and free-contact allowance that directly conflicted with figures from separate research conducted the same week. Not a rounding difference. Different numbers, for the same product, on the same date, one of them from ManyChat's own material.

Given that, a dated historical timeline would have been a chain of guesses presented as research. Every entry would have inherited the uncertainty of the entry before it, and we'd have dressed it up with dates to make it look rigorous. So the timeline is the one thing this article does not contain.

A number you can't source is not data. It's a rumour with a decimal point.

Check it yourself — here's the method we used

Rather than ask you to trust our failure, here's how to run the check independently. It takes about ten minutes and it works for any vendor, including us.

1. 1Open the vendor's live pricing page and screenshot it, with the date visible. This is your primary source and it will not be true forever.
2. 2Run that URL through the Internet Archive's Wayback Machine, which holds dated snapshots of most pricing pages going back years. This is the only genuinely neutral pricing history available — it's captured mechanically, not published by anyone with a stake.
3. 3Compare snapshots a year apart. Look at the tier boundaries — the contact counts — as carefully as the prices, because a tier boundary moving down is a price rise that leaves the price alone.
4. 4Check your own invoices. Your billing history is the only pricing record that is definitionally true for you, and it beats every published tier table.

We're pointing you at the archive rather than summarising what we found in it deliberately. If the numbers are contested, you should see the primary source rather than our reading of it. A competitor summarising a competitor's price history is exactly the kind of source you should distrust — including when the competitor is us, and including when the summary happens to be accurate.

The pricing change nobody announces

Here is the thing the timeline would have missed entirely, and it's the change that actually hits most accounts.

ManyChat bills per contact, in bands. Your bill therefore rises without the price list changing at all — you simply cross a boundary. No announcement, no email, no entry on anybody's timeline. The mechanics are in our breakdown of per-contact billing, but the point here is about visibility: this is a price increase that is invisible to exactly the kind of article you're reading.

Which reframes the search. If you're here because your bill went up, the overwhelmingly likely cause is not a pricing change. It's a good marketing month. A reel overperformed, a giveaway ran, a comment-to-DM keyword caught — and the contacts those wins collected pushed you across a band. Your price changed because your audience did.

• It's not announced, because nothing on their side changed.
• It's not reversible without work — contacts don't expire automatically, so you're pruning a list by hand to get back under a line.
• It scales with your success, which means it recurs exactly when things are going well.

Three more ways a bill rises without a price rise

Band-climbing is the common one, but it isn't the only mechanic that makes a timeline of headline prices misleading. These are the four to check on any vendor — we're listing them as things to verify in your own account, not as accusations about anyone's:

• The boundary moves, not the price. If a tier that used to cover twice as many contacts quietly covers half as many, the advertised price is untouched and everyone who falls into the gap just got a rise. This is why you compare tier boundaries between archive snapshots, not only the dollar figures — and it's why we're describing the mechanic in ratios rather than putting numbers on it.
• Something you already use moves behind a higher tier or an add-on. Unbundling is a price rise wearing a feature announcement's clothes. The tier table looks stable; your configuration costs more.
• Grandfathering ends. Legacy plans get honoured for years and then don't. Nothing on the public pricing page ever reflected your old plan, so nothing on a public timeline can show it disappearing.
• The promo you signed up on lapses. Your first-year price was never the price. This one shows up in your invoices and nowhere else.

Every one of those is invisible to the article you thought you were opening. A timeline of published headline prices could be perfectly accurate and still tell you nothing about what happened to your bill — which is a decent argument that the timeline format is broken for SaaS generally, not just here.

Why every comparison site quotes the same number

You'll notice something if you search this topic: a lot of sites state ManyChat's pricing with total confidence, and many of them agree with each other. That agreement is worth exactly nothing, and understanding why will make you a better reader of every comparison article you meet.

They agree because they copied each other. One post quotes a figure, the next cites the first, and within a year a number nobody re-checked has become consensus through repetition alone. Multiple sources saying the same thing feels like corroboration; here it's an echo. The tell is that they rarely name the date they checked, or link to a snapshot you could verify — because checking is work and copying is free.

We found this out the hard way: our own earlier writing on ManyChat's pricing quoted band figures with a hedge that the exact numbers shift. We stand by the hedge and we'd write it more loudly now. If a figure matters to your decision, the only sources that count are the vendor's live page, a dated archive snapshot, and your own invoice.

The second line that moves with it

The conversational AI features sit behind an add-on stacked on the base plan, so the same growth that pushes your contact tier up also grows the line riding on top of it. Two bills, one cause. If you're reconstructing why an invoice jumped, check both.

Credit where it's due

The most interesting thing we found in ManyChat's own 2026 FAQ was not a price. It was that the FAQ candidly suggests users can skip the AI add-on and wire up their own OpenAI setup instead.

That is a vendor publishing instructions for spending less money with them. Whatever else is muddy about the pricing, that is not the behaviour of a company trying to trap anyone, and a fair article says so plainly. It also tells you something structural, if you're buying: the AI is positioned as a convenience layer over an API you could call yourself, rather than as the engine of the product.

Our pricing changes too

It would be convenient to end here, having established that a competitor's prices are hard to track. So, symmetrically: ours change as well. We're a young company, and any vendor in this category who promises you a price frozen forever is writing a cheque their next funding round will bounce.

Run the same four checks on us. Archive our pricing page today so you have a dated snapshot that doesn't depend on our word. Watch whether our tier boundaries move. Watch whether anything you already use migrates behind a higher plan. If we ever do any of those quietly, the method above will catch us, and it should.

We'd also rather you noticed the asymmetry in this article than have us point at it: everything critical here is about a mechanic, not a motive. We have no evidence ManyChat has done anything underhanded with pricing, and we've said several times that their own FAQ points users toward the cheaper option. What we're arguing is narrower — that per-contact billing produces bill rises nobody has to announce, and that's a property of the model rather than a decision anyone made about you.

What we'd rather be judged on is the unit. HyperDM bills per conversation — you pay when the agent actually replies and does work — so the specific failure mode described above does not exist here: there is no contact band to cross, and 40,000 silent new followers cost nothing. That is a structural claim about the meter, not a promise about the number, and you should hold us to the difference. The current numbers are on the pricing page, the free tier is 50 conversations a month on one channel with no card, and you can model your own volume on the cost calculator.

The real story behind ManyChat pricing changes

The honest summary: we can't give you a timeline of ManyChat pricing changes, we don't believe anyone else can either, and the confident-looking tables you'll find elsewhere are almost certainly reporting numbers with more precision than the source supports. Check the archive and your own invoices.

More importantly, the timeline was the wrong question. For most accounts the bill rises because the contact count crossed a line, not because a price did — which means the fix isn't waiting for better pricing, it's deciding whether you want a meter that counts your audience at all. Whether that's worth switching over depends on your own numbers, and we worked through that arithmetic in is ManyChat worth it. If your answer is no, the switch page has the move.