Free

What is your AI actually costing you right now?

Install Recovea beside the provider client you already use. Seven days of reporting later, your own numbers: by agent or workload if you run your own, by client or customer if they pay for what your AI does.

Start your Baseline

One number, or the split underneath it.

Illustration

contract-review
support-triage
nightly-reindex
research-agent
sales-drafts

One line on one invoice, and nothing behind it.

What the Baseline tells you.

The headlineYour reported spend for the seven days and what it annualizes to, with the gap against your last invoice total if you declare one.
The worst hourThe most expensive hour in the seven days, as a multiple of your median hour.
Attribution by unitEvery reported call attributed to the unit it carried, busiest first.
ConcentrationThe models and units that carry most of the reported calls, ranked.
Loop and retry signatureRuns of near-identical calls at a fixed cadence, retry storms and fan-out spikes, found in the envelopes and never in what the calls said.
Cost per requestCost per request, with no success signal: on this wire the absence is the finding.
The exposure ceilingA modeled bound: what one runaway loop at your own peak rate would cost over twenty-four hours, beside whatever bounds it today.
Three closed doorsYour per-client or per-agent breakdown with its largest row already open, exposure by agent or workload, and the five moves ranked as bounds.

It runs for seven calendar days, counted from the first whole day your application reports priced calls, and then closes. The page counts the days, never the calls.

Ten minutes, three steps.

1
Install the package
JavaScript
```
npm i recovea
```
Python
```
pip install recovea
```
2
Put your key in the environment
Shell
```
export RECOVEA_TAP_KEY="your key"
```

Wrap the client you already use

JavaScript

import OpenAI from "openai";
import { tap } from "recovea";

const openai = tap(new OpenAI(), { tag: "matter-4821" });
// use openai exactly as before

Python

from openai import OpenAI
from recovea import tap

client = tap(OpenAI(), tag="matter-4821")

Your calls reach your provider exactly as before. The line reports the envelope after the call has returned, so nothing of ours is ever on your path.

Install Recovea beside the provider client you already use. Nothing routes through Recovea, and the line never reads a prompt, a reply or a provider key.

Computed from the events your own application reported. Export them and check the arithmetic.

See your own numbers.

Start your Baseline