The shape of it
One month cost half of everything
Here is the whole history. April is not a trend, it is an accident —
and an instructive one, because nobody noticed it while it was happening.
Mar$26.18
Apr$309.89
May$75.86
Jun$61.13
Jul$86.67
Aug$78.88
Sep*$10.08
* 1–18 September, partial month. Bars are scaled to April. Every figure is
the sum of the usd field across the usage ledger for that month.
What a blowout looks like from the inside
April was $309.89 across 23,065 calls — 48% of six months' spend in
four weeks. Two things caused it. The system was routed to a reasoning model priced at
$30 per million in, $180 per million out, and it was running multi-agent
strategy panels that fan a single question out to several expensive models at once. 1,040
calls to that one model cost $93.51 by themselves.
Nothing was broken. No alert fired, because no cap was breached — the global daily
ceiling was never close to binding. The bill was simply the honest price of a routing
decision nobody had costed, made once and left in place for a month.
Same model class, a generation later, is a sixth of that price. Most of what
looks like optimisation above is just the market moving underneath us — worth saying
out loud, because it is not a skill we can claim.
The actual finding
The waste was not the model price. It was the retries.
Every cost conversation about agents is about picking a cheaper model.
Ours turned out to be about something duller: a fallback chain full of rungs that could never
succeed, retried thousands of times, each attempt billed.
Failed calls — 14,069 of 31,755, at $108.11
| Rung | Failed calls | Cost | Why it could not succeed |
|---|
| A frontier model on a dead account | 4,453 | $54.16 | API credits exhausted; the rung stayed in the chain |
| A general workhorse | 4,546 | $27.46 | genuine transient errors and overloads |
| A preview model with no key | 4,899 | $23.85 | never authenticated, not once, in six months |
| Three rungs | 13,898 | $105.47 | 98.8% of all failures |
The third row is the one that stings. A model the system had no credentials for
sat in the fallback chain for six months and was tried 4,899 times. Its success rate was
effectively zero. It was not a bug anyone could see — the chain worked, because the
next rung answered every time.
Why a broken rung is invisible
A fallback chain is designed to hide exactly this. Rung one fails, rung two answers, the
caller gets its result, and nothing upstream can tell the difference between “first
choice answered” and “first choice has been dead since March.” The system
reports healthy. It is healthy. It is also paying a toll on every request, and
adding a second or two of latency to every task, forever.
We only found it by parsing the usage ledger by status rather than by model — which is
not a view any dashboard offers by default, because the interesting number is a ratio between
two columns nobody thinks to divide.
Read this before quoting the number
The daily rate is not pure efficiency
September's $10.08 over eighteen days works out to about 56 cents a day, and
it would be dishonest to present that as pure efficiency. Much of the agent's reasoning moved
onto a subscription-authenticated path that this ledger does not bill at all. Some
capability got cheaper. Some capability simply stopped appearing on this particular invoice.
The ledger has other blind spots we know about: a dozen scripts that call models directly
instead of through the router, all gateway traffic, and any model whose price the system
doesn't recognise, which logs as $0.00 rather than as unknown. A cost figure with unlisted
exclusions is a marketing number. These are ours.
What transfers
Four rules, if you are running agents on a budget
Rule 01
Audit by status first
Cost per successful result is the only figure that means anything. Ours was 20%
higher than the sticker price on every model we used, and no per-model view would ever have
shown that.
Rule 02
An unauthenticated rung is not a fallback
It is a scheduled payment. Delete guaranteed-fail rungs the day they start failing, and
make the chain report which rung answered, not just that something did.
Rule 03
Price the routing, not the model
The April blowout was one configuration change nobody costed. Caps didn't catch it because
caps guard against runaway volume, and this was ordinary volume at six times the unit
price.
Rule 04
Cut waste, never capability
Every saving above came from deleting calls that produced nothing, or from prices falling.
Not one came from giving the agent a weaker model for the same job. A cheaper agent that
makes worse decisions is not a saving, it is a slower way to lose money.
The offer
What the $648.69 bought
Six months of an agent running a business end to end — and, in the last three weeks,
1,146 cold emails for a paying-client campaign that produced its first held meeting. That story
is published with the same discipline as this one: full denominator, failures included.
Rick Acquisition — $800/month
Figures drawn from the agent's usage ledger on 18 September 2026 — 31,755
rows, every one parsed, nothing sampled. Model names are generalised because the specific
versions date faster than the lesson does. Prices quoted are those in effect when the calls were
billed and are not current vendor pricing. No model vendor is affiliated with or has endorsed
this page.