The subscription is a convenience, not a gate
Yes — every AI feature in MindMax works with your own API key, with no account, no weekly allowance and nothing held back. The subscription exists to save you from managing keys and invoices, not to unlock the product. This post explains why, and what running on your own key costs you.
If you already pay for an API, being told to buy a second subscription before an app will talk to a model is a bad trade, and everybody involved knows it. You are being charged twice for the same tokens — once by the provider, once by the app for the privilege of sending them.
So MindMax does not do that. Add your own key and every AI feature works the same way it does for a subscriber: the same tiers, the same tools, no account, no weekly allowance. Not a reduced mode, not a beta path, not a setting that quietly disables the interesting half of the product.
This post is about why that is not generosity.
How the gate grows
The gate is not usually put there out of greed. It grows out of an honest constraint: if an app pays for model usage on your behalf, then model usage is its cost of goods. Every long job you run is a real invoice arriving at somebody's office. Under that structure, the ability to bring your own key is dangerous — it lets the heaviest users, the ones a usage-based plan was priced around, opt out of paying while still consuming everything else.
The defensive move is to make your own key a second-class path. Support it, but not for the expensive features. Support it, but require an account anyway. Support it, but never mention it on the pricing page.
We looked at that and concluded it buys very little and costs a lot.
What the arithmetic actually says
The people most likely to want their own key are the people who already have one: they have an API console open, they know what a token costs, and they have opinions about which model is good at what. That group is not deciding between our subscription and our free tier. They are deciding whether to install anything at all.
Locking features away from them does not convert them. It loses the download. And a download from that group is worth an unusual amount, because they are the ones who find the sharp edges, who compare a reading pass across three models, and who tell other people about tools.
So the subscription has to earn its place a different way. What it sells is not access — it is not managing four keys, not reconciling four invoices, not discovering that a model was retired when a job fails at eleven at night. That is a real product, and it is one we can describe out loud without flinching. "Pay us or the software does not work" is also a product, but it is not one that survives the first person who does the arithmetic.
The place where you can check whether we mean it
Marketing copy is cheap. The interesting evidence is in the parts of the codebase nobody was meant to read.
MindMax can look things up online during a conversation. On the plans that include it, and with nothing of your own configured, that goes through our gateway and spends credits. But if you have your own search provider key configured, the gateway provider steps aside — it reports itself as unavailable so your own key is used instead. The comment sitting above that decision has been in the repository since long before this post, and it says, in as many words: the subscription's selling point is that it works without configuring a key, not that we will pay for something you have already paid for.
Nobody would have noticed if it worked the other way round. Routing your searches through our gateway would have shown up as engagement, not as a bill you are paying twice. That is the kind of decision that tells you what the defaults will look like in two years.
What it costs you
The honest version has four parts, and none of them are small.
No allowance means no brake. The weekly credit limit is what stops a subscription running away with itself. With your own key that limit does not exist here — a job that reads a whole book costs what it costs, and the invoice is between you and your provider. The usage page counts tokens so it is at least visible, but nothing stops you.
Not every provider can do every job. Reading a scanned page or a boxed region of a chart needs a model that takes pictures. A text-only provider leaves that tier empty, and the features that depend on it have nothing to run on until you add a second provider for that one tier.
A model on your own machine will not carry the long work. Short questions about an open document are genuinely fine. Reading a book chapter by chapter, translating three hundred pages, drafting something book-length — those lean on long context and dependable tool calls, and a model sized to a laptop will be slow at them or will not finish them. We would rather say this than let you discover it four hours in.
Maintenance becomes yours. Keys expire, providers retire models, endpoints move. On the subscription that is our problem. On your own key it arrives as a job that suddenly fails.
One thing that is not true
It would be wrong to read any of this as "bring your own key, and your documents stay on your machine". Those are different claims and only one of them follows.
Pointing a tier at a cloud provider means the text goes to that provider rather than through us. That is a meaningful difference — it is one less party — but it is not the same as nothing leaving your machine. The only configuration where that is true is a model running locally, and we say so in that specific spot rather than letting the general claim drift over the whole feature.
We would rather be believed about the small claim than admired for the big one.
The practical version is on bringing your own key, including which providers come pre-filled and where the awkward parts are. Setting one up is a short walkthrough, troubleshooting included. And if you would rather not manage keys at all, that is what the subscription is for — which is the whole argument, really.