EquipTalkIQ

AI · Usage and budgets

Know what the AI costs, call by call, before the month ends.

Every model call is recorded with who asked, which feature, the model, tokens, the cost at list price and the records it touched. Admins see the breakdown, set a monthly budget with a warning and a hard stop, and export it all.

Available today
EquipTalkIQ AI usage and budgets screen with sample records
AI usage · total cost, calls, tokens and the budget, with spend by featureApp screen · sample data

What changes

Today

  • AI cost arrives as one number on someone else’s invoice.
  • Nobody can say which feature, person or project drove it.
  • An overrun shows up after the money is spent.

With EquipTalkIQ

  • Every call is a row: person, feature, model, tokens, cost and records touched.
  • Spend breaks down by feature, model, person, billing party, site and project.
  • A monthly budget warns near the limit and can stop new calls at it.

How it works

4 steps, in the app as it ships.

  1. Every call is recorded

    AI usage · Total cost, Model calls, Tokens, People

    Each request to the model writes one record: who asked and their account type, the product feature, the model, tokens in and out, the cost at that model’s list price, whether it succeeded, failed or was blocked, and the site, project and conversation it belonged to.

  2. Break it down

    By feature · By model · By person · By site · By project

    Pick a period and filter by feature, model, person or billing party. Click a bar to narrow everything to one feature. Most expensive runs lists the costliest conversations and background document runs, each with a link to open it.

  3. Set the monthly budget

    Monthly budget · Warn at (% of limit) · Hard stop at the limit

    Owners and admins set a monthly limit in dollars and the point to warn at. The warning shows under the AI chat. With the hard stop on, a call at the limit is refused before any tokens are spent and counted as blocked.

  4. Export it

    Export .xlsx

    Download a spreadsheet with the summary breakdowns and one line per call, including its status and the records it touched.

Receipts

What you can check for yourself.

Priced at list, and dated

Cost uses the published list price of the model that answered, and the card says which price list applied.

The Total cost card

Blocked calls are counted

Calls refused by the budget and calls that failed are counted beside the rest.

The Model calls card

Invited guests shown apart

Questions from invited owners and reps are billed to your company and shown under their own billing party.

By billing party · Invited guests

The warning reaches the person asking

Near the limit, the chat shows the month’s spend against the budget; at the limit it says new requests are paused.

The note under the AI chat

One line per call, exported

The spreadsheet holds every call with person, feature, model, tokens, cost, status, site, project and records.

Export .xlsx · Calls sheet

What it won’t do

  • Charge you automatically for going over: nothing is billed without you choosing it.
  • Let a call through at the limit while the hard stop is on.
  • Show the cost breakdown to everyone: it is for company owners and admins.

Who it’s for

The people on this screen.

  • Company owners
  • Admins

For mechanical contractors

Works with

  • Ask with sourcesIn the app

    Every unit, site and portfolio question is a metered call.

  • Specs, takeoff and draftsIn the app

    Spec indexing and logs, takeoff runs, plans, claims and briefs are metered the same way.

  • Owner and rep assistantsIn the app

    Guest questions are billed to the company whose data they asked about.

All connections

Questions

Worth asking.

How is a call priced?

At the published list price of the model that answered, for the tokens it used. The price list is versioned, so the page shows which one applied.

What happens at the budget limit?

With the hard stop on, which is the default, new calls are refused before any tokens are spent and recorded as blocked. People asking see that the month’s budget is reached.

Are we charged automatically if we go over?

No. Your plan includes an AI allowance sized for normal use. If you need more, you choose a higher allowance or a fixed-price pack; nothing is charged automatically.

Who can see the usage page?

Company owners and admins. They are also the only people who can change the budget.

Do invited owners and reps cost us?

Their questions are billed to the company whose building or rooms they asked about, and shown under Invited guests so you can see them separately.

Keep exploring

See it on your own work.

We load your sites, units and documents during the pilot, so the walkthrough is on your own kind of work. The 90-day pilot is $9,000, credited to your first year if you continue.