And This Is the Best Part.

Enterprise-grade municipal AI shouldn’t require enterprise drama.

A single department typically spends $60k–$100k per year answering routine policy and service questions. Agency Chat costs less than one part-time staff role — and can be live in weeks.

Base Subscription

$1,495/month

Fully functional site that shows its work.

No multi-year lock-ins
No speculative AI training cycles
No per-seat licensing
No hidden onboarding fees
No hidden implementation fees
No mandatory 3-year contracts
No custom pricing negotiations
No vendor negotiation theater

Included each month

10,000

AI interactions

5M

DeepL translation characters

Prepaid and overage pricing

AI interactions

Prepaid: $17.50/1,000

Overage: $25/1,000

DeepL translation characters

Prepaid: $30/1M

Overage: $40/1M

Usage is metered monthly by successful AI interactions and translated characters. Prepaid credits are consumed before overages, and overages are billed in arrears.

Contact us

Department-level activation · No login required · Cancel anytime

Billing & Metering FAQ

How is usage measured?

AI usage is measured by successful interactions. Translation usage is measured by translated characters.

What gets consumed first each month?

Included monthly usage is consumed first, then prepaid credits, then overage.

When are overages billed?

Overages are billed monthly in arrears and appear as separate invoice line items.

Do unused included amounts roll over?

No. Included monthly allowances reset each billing cycle and do not roll over.

Do prepaid credits expire?

Unless otherwise stated in your agreement, prepaid credits are valid for up to 12 months.

30-day pilot

Start small. Prove value. Walk away if it doesn’t.

Every deployment begins with a live 30-day evaluation against real documents and real residents — with a clean exit ramp if it doesn’t earn its keep.

What’s included

  • A fully-branded deployment on your subdomain
  • Up to 250 source documents indexed
  • Recurring crawl of your public website
  • 72+ language coverage and full accessibility
  • Public transparency dashboard, live from day one
  • Direct support from the team that built it

What gets measured

  • Real resident traffic and question categories
  • Citation rate and refusal rate per category
  • Translation use across non-English languages
  • Peak-hour volume for capacity planning
  • A first CPRA/FOIA-ready records report at day 30

Exit ramp

  • Month-to-month after the pilot — cancel any time
  • No multi-year contracts, no per-seat licensing
  • Final records report delivered on offboarding
  • Your documents and indexed corpus stay in your GCP project
  • No proprietary data layer to hold you hostage

What you’re not paying for, ever

Implementation consultants
Speculative AI training cycles
Per-seat user licenses
Hidden onboarding fees
Feature-gated plans
Mandatory 3-year contracts
Custom development add-ons
Vendor negotiation theater
Start a 30-day pilot

Typically live in 1–2 weeks · First records report at day 30