Pricing calculator

Estimate your bill and your payback

Move the sliders to your team, your project, and how often people ask. The bill follows the published rate card, per completed agent run. The return follows assumptions you can change.

Get a Demo See the rate card

Typical return

20–50×

Engineering time saved against the SignalPilot bill. Model tokens are shown separately; you would pay them for any agent.

Per successful thread

From $0.50

Plus $0.01 per governed query it ran. Failed threads, refused queries, and answers below the SLA-backed evidence percentage are never billed.

Wrong numbers avoided

Priceless

Double-counted joins, missing refunds, last quarter's definitions in this quarter's deck.

Your team

Members with a seat50

Everyone who should be able to ask a question. Plans include 5, 15, or 100 seats; Enterprise scales to 2,000.

Data models in the project200

Sets the tier and the coverage allowance.

Share of models with evals80%

Models with at least one eval. The included allowance comes off this.

Usage

A thread is one full agent run: the question, the plan, every governed query it needs, the checks, and a receipted answer.

Successful threads per member, per month20

One full agent run that ended in a receipted answer. Runs that fail are free.

Governed queries per thread30

Typical is 20 to 40.

Eval runs beyond the 30 included20

Pull request and on-demand runs. About 40 queries each.

Model class

Managed plan

Time saved

Hours saved per successful analysis2 h

Writing the query, checking the join, reconciling, sharing. Conservative default.

Share of threads that would have gone to the data team75%

The rest are questions nobody would have asked.

Loaded hourly cost of the engineering resource$100

Context and eval upkeep hours avoided, per month16 h

8 h per 100 data models, rounded up. Halved on Managed.

Your plan

Scale

50 seats is above 5, within the 100 Scale covers.

$2,153

per month · $25,836 a year

Plan fee, seats included

$250 includes 15 seats, plus 35 added at $15

$775

Completed runs

1,000 threads at $0.50

$500

Governed queries

30,000 queries at $0.01

$300

Extra eval runs

20 runs at $0.50 plus queries

$18

Included usage credit

$100 a month on Scale, applied to usage

−$100

Data models with evals

160 covered, 50 included, 110 at $6

$660

Managed

Off

SignalPilot, per month$2,153

Model tokens, billed separately

1,000 threads on a Sonnet-class model, passed through at your provider's list price with 0% markup, or bring your own key. You would pay this for any agent, so it is not counted in the return below.

$400

Return

1,516 h

saved per month

$151,600

of engineering time, per month

70×

return on SignalPilot cost

Cost per completed agent run

One run is a question, its plan, about 30 governed queries, the checks, and the evidence graph. Model tokens separate, about $0.40 a run at cost.

$2.15

Engineering hour, for comparison

$100

Net value per month

Engineering time saved minus SignalPilot cost

$149,447

Every number above is an estimate from the assumptions on the left and the published rate card. Failed threads, refused queries, and threads below the SLA-backed evidence percentage are never billed. See the full rate card →

Start with one workflow.

We will run the first workflow on your stack so the estimate becomes a number.

Get a Demo Talk to Us