Seats vs. a shared API key: do the arithmetic before you buy
Most teams overpay for AI by defaulting to per-seat plans. Here's the napkin math that tells you when a shared API key is cheaper, and when it isn't.
Every vendor sells AI the same way: a price per person, per month. It's a clean number, it's easy to expense, and for a lot of teams it is the wrong way to buy.
The alternative is boring: one API key, billed by usage, wrapped in something your team can actually use. It's less convenient. It is often dramatically cheaper. The only way to know which side you're on is to do the arithmetic, so let's do it.
The whole model in four lines
A message costs a few fractions of a cent. With Sonnet at $3/M input and
$15/M output tokens, a typical message (call it 2,000 tokens in, 700
out) costs about $0.0165. That's it. That's the unit.
From there:
- Seats cost
people × seat price. Ten people on a$30plan is$300/mo, full stop, whether they send one message or a thousand. - A shared key costs
people × messages/day × working days × cost/message. Those same ten people, sending15messages a day across21working days, run about$52/mo.
At light usage it isn't close. The seat plan costs almost six times as much for the same work.
That 2,000 tokens in, 700 out isn't an abstraction; it's a real amount of
text. Drag the sliders and see for yourself how much one message actually is:
How much is a token, really?
Watch how much real text one AI message actually is.
input 63% · output 38%
Alice was beginning to get very tired of sitting by her sister on the bank, and of having nothing to do: once or twice she had peeped into the book her sister was reading, but it had no pictures or conversations in it, "and what is the use of a book," thought Alice, "without pictures or conversations?" So she was considering in her own mind, as well as she could, for the hot day made her feel very sleepy and stupid, whether the pleasure of making a daisy-chain would be worth the trouble of getting up and picking the daisies, when suddenly a White Rabbit with pink eyes ran close by her. There was nothing so very remarkable in that; nor did Alice think it so very much out of the way to hear the Rabbit say to itself, "Oh dear! Oh dear! I shall be late!" But when the Rabbit actually took a watch out of its waistcoat-pocket, and looked at it, and then hurried on, Alice started to her feet, for it flashed across her mind that she had never before seen a rabbit with either a waistcoat-pocket, or a watch to take out of it, and burning with curiosity, she ran across the field after it, and fortunately was just in time to see it pop down a large rabbit-hole under the hedge. In another moment down went Alice after it, never once considering how in the world she was to get out again. The rabbit-hole went straight on like a tunnel for some way, and then dipped suddenly down, so suddenly that Alice had not a moment to think about stopping herself before she found herself falling down a very deep well. Either the well was very deep, or she fell very slowly, for she had plenty of time as she went down to look about her, and to wonder what was going to happen next. First, she tried to look down and make out what she was coming to, but it was too dark to see anything; then she looked at the sides of the well, and noticed that they were filled with cupboards and book-shelves: here and there she saw maps and pictures hung upon pegs. She took down a jar from one of the shelves as she passed; it was labelled "ORANGE MARMALADE", but to her great disappointment it was empty: she did not like to drop the jar for fear of killing somebody underneath, so managed to put it into one of the cupboards as she fell past it.
Alice's Adventures in Wonderland, Lewis Carroll (1865). Public domain, Ch. 1.
Breakeven is the only number that matters
The honest question isn't "which is cheaper?" It's "how heavily do we use it?" There's a single message count where the two prices meet:
breakeven = seat price ÷ (working days × cost per message)
For the numbers above that's about 87 messages per person per day. Below that,
the API key wins. Above it, seats win, and you stop worrying about overages.
Eighty-seven messages a day is a lot. Most teams never get close, which is why
the default plan quietly overcharges them.
The catch nobody prices in
A raw API key is not a product. Somebody has to put a login, an interface, and
some guardrails on top of it before your team can use it. That's real work, or a
wrapper tool you pay for separately. The $52 is the ingredient cost, not the
finished meal.
So the rule is simple. Estimate your real messages-per-day, find your breakeven, and only chase the API savings if you're comfortably under it and you have something to put the key behind. If you're a heavy-usage team, stop reading and buy the seats; you've already won.
Don't take my numbers. Use yours. Put your team's size and usage in here:
Seats vs. a shared API key
Put in your team's numbers and see which way of buying AI is cheaper.
Use a shared API key.Saves $248 / month, about 83% less than seats.
$0.0165 / message
$1 of Claude Sonnet 4.6 ≈ 151,515 tokens ≈ ~227 pages of reading. what's a token?
A raw API key isn't a product. Someone still has to put an interface, login, and guardrails on top. Budget for that build (or a wrapper tool) before treating these savings as free.
A raw API key still needs an interface, login, and guardrails on top; budget for that before treating the savings as free.