AI CHAT CONCIERGE

// PRICING · USD · NO PER-SEAT GOTCHAS

You pay for the platform. The AI is one component.

Elasticsearch-indexed knowledge base. Multi-channel routing. Operator apps. Analytics. Escalation. Booking. That's the product. Cheap models included in every tier — for Sonnet / Opus / GPT-4 add BYO API key +$50/mo and route through your own Anthropic or OpenAI account.

01
Four tiers
Basic · Starter · Growth · Enterprise — each additive.
02
What counts as a message
Only AI replies to customers.
03
Bring your own API key
+$50/mo — any Anthropic or OpenAI model, unlimited.
04
What we actually charge for
Elastic KB, routing, ops app, analytics — not the AI.
05
FAQ
Which models? Overages? BYO? Annual?
06
Enterprise + white-label
Chains, franchises, agencies.
Regional deployment on requestGDPR-ready DPA on requestiOS + Android appData export anytimeFace ID gateElasticsearch-indexed KB

// 01 — FOUR TIERS · ADDITIVE

Each tier includes everything from the tier below.

ENTRY

Basic

$49 / month

The AI concierge on your website. One widget, one team seat, everything you need to launch.

  • Website widget on 1 site
  • WordPress + Shopify plugins
  • 1,000 AI messages / month*
  • 1 operator seat
  • Knowledge base — unlimited docs
  • Booking · Google Cal / M365 / Cal.com / Calendly
  • Text replies · 40+ languages
  • Email support · 24h reply
  • Setup guided · live in 1 business day

ADD A BRIDGE

Starter

$149 / month

Everything in Basic, plus 1 dedicated bridge — pick WhatsApp, Telegram, OR Signal.

  • Everything in Basic, plus:
  • 1 bridge slot · pick WhatsApp / Telegram / Signal
  • 3,000 AI messages / month*
  • 3 operator seats
  • Image reading (send a photo, AI describes / answers)
  • Scheduled announcements
  • Operator hand-off · Slack + Teams + email
  • Email support · 24h reply

MOST POPULAR

Growth

$499 / month

Everything in Starter, plus 3 bridge slots — enable WhatsApp + Telegram + Signal all at once.

  • Everything in Starter, plus:
  • 3 bridge slots · WhatsApp + Telegram + Signal simultaneously
  • 15,000 AI messages / month*
  • 15 operator seats
  • Voice notes (transcribe + reply)
  • Host images + documents in KB
  • MCP integrations · custom bots
  • Analytics dashboard
  • Custom operator-app domain
  • Email + WhatsApp support · 4h reply

CHAINS · WHITE-LABEL

Enterprise

Custom

Everything in Growth, plus unlimited bridges, Email channel, white-label, SSO, dedicated support.

  • Everything in Growth, plus:
  • Unlimited bridges
  • Email channel · dedicated inbox
  • White-label · custom domain
  • Unlimited AI messages
  • Unlimited operator seats
  • SSO (SAML / OIDC)
  • MCP developer support
  • Regional deployment on request
  • BAA · DPA · custom SLA
  • Dedicated Slack channel · 1h response

// * BUNDLED MODELS

What runs on us.

AI

Included models · fast, cheap, accurate for 90% of chats

Claude Haiku 4.5 · GPT-4o-mini. These are the models we run for you on every bundled tier — plenty of intelligence for FAQ answers, booking flows, order lookups, escalation triage, and 40+ language support. Pick Anthropic or OpenAI per instance.

>>

Want Sonnet, Opus, GPT-4, or Fable? Add BYO.

Frontier models are on a Bring Your Own API Key add-on (see below). You paste your Anthropic or OpenAI key, we route every reply through it, your model bill lands directly in your provider account — no black-box markup from us.

$.08

Overage · $0.08 per message

Every AI reply above your monthly cap costs 8¢. Alerts fire at 80%, 100%, and 120% of the cap. Upgrade mid-cycle for a prorated fee if you're consistently past the ceiling. No surprise cutoffs.

// 02 — WHAT COUNTS AS A MESSAGE

Transparent. Because surprise bills kill trust.

YES

What counts

Every AI reply to a customer counts as one message. Scheduled announcements count as one per recipient. Image reading (Starter+) and voice-note transcription (Growth+) count the same as a text reply.

NO

What doesn't

Operator-sent messages, operator-dashboard actions, customer messages that never trigger a reply (spam, empty, greetings we skip), retries or edits after an operator takes over. KB indexing, storage, and search — free, no matter how many docs you upload.

80%

How we alert

Email + in-app notice at 80%, 100%, and 120% of your monthly cap. Overages billed at $0.08 per message (this covers cheap-model inference + platform + margin — no hidden mark-ups), or you can upgrade mid-cycle for a prorated fee. No surprise cutoffs — service keeps running.

// 03 — BRING YOUR OWN API KEY

Flat +$50 / month. Any tier. Any Claude or GPT model.

+$50

Flat add-on on any bundled tier

Basic + BYO = $99/mo. Starter + BYO = $199/mo. Growth + BYO = $549/mo. Enterprise BYO — custom. Same platform, unlocks any Claude or GPT model plus removes the monthly message cap (fair-use throttling only).

HOW

Pick provider · pick model · paste key

In your dashboard, choose the provider (Anthropic or OpenAI), pick the model (Sonnet 4, Opus 4, Fable 5, GPT-4o, GPT-4.1, etc.), and paste the API key. Every AI reply is routed through your account with that model. Per-instance setting — a Growth account can run Sonnet on the widget and Opus on WhatsApp.

WHY

You keep control of the inference bill

Your model bill lands directly in your Anthropic or OpenAI console at their published wholesale rate — no black-box markup, no per-message surcharge, no surprise line item on our invoice. Spend existing Anthropic enterprise credit or OpenAI free-tier quota first. Full audit log of every routed call.

=

That's the honest deal

Elasticsearch-indexed KB, routing, ops app, analytics, escalation, booking — that's the platform. Cheap-model inference is basically free at our scale. Frontier-model inference is not, so we don't underwrite it. You want Sonnet, you pay Anthropic. We take $50 for the convenience of running it in our platform.

Unlimited KB documentsData export as JSON, anytimeHTTPS · TLS 1.3Regional deployment on requestGDPR-ready DPA on requestiOS + Android + web ops appFace ID / Touch ID gateEvery new feature at your tierNo per-seat fees inside your capElasticsearch-indexed KBMulti-channel routing · web + WhatsApp + SMS + Telegram + Signal + EmailEscalation to Slack + Teams + emailBooking · Google Cal + M365 + Cal.com + Calendly

Pricing FAQ.

Which AI models are included in bundled tiers?

Claude Haiku 4.5 · GPT-4o-mini. These are Anthropic's and OpenAI's fast + cheap models — plenty for FAQ answers, booking flows, escalation triage, and 40+ language support. If you want Sonnet, Opus, Fable, GPT-4o, or GPT-4.1, add BYO API key (+$50/mo) and bring your own Anthropic or OpenAI account.

Do you charge per user?

No. Every tier has an operator cap (Basic 1 · Starter 3 · Growth 15 · Enterprise unlimited). Anyone under that cap is free — add or remove seats freely without a call to sales.

What happens if I exceed my message limit?

We alert you at 80%, 100%, and 120% of the monthly cap. Overages billed at $0.08 per message, or you can upgrade mid-cycle for a prorated fee. No surprise cutoffs — service keeps running. Consistently past the cap? Add BYO (+$50/mo) and the cap goes away (fair use only).

How does BYO API key work?

Add-on is a flat +$50/mo on any tier. In your dashboard you pick the provider (Anthropic or OpenAI), the model (Sonnet 4, Opus 4, Fable 5, GPT-4o, GPT-4.1, etc.), and paste the API key. We route every AI reply through your account with that model — model + key is per-instance, so a Growth account can run Sonnet on the widget and Opus on WhatsApp. Fair-use throttling to prevent abuse; your Anthropic / OpenAI bill lands directly in your provider console at their published rate.

Why only Anthropic and OpenAI for BYO?

They are the two providers with production-grade tool-use, native prompt caching, and reliable model families across cost tiers (cheap → frontier). If you need Google Gemini, Mistral, or a self-hosted model, ask us — Enterprise tier can accommodate.

Can I upgrade / downgrade mid-cycle?

Yes. Upgrades take effect immediately with prorated billing. Downgrades take effect at the end of the current cycle so you keep everything you already paid for. Cancel anytime — 30 days of data export after cancellation.

Do you offer nonprofit / education pricing?

Yes — 50% off Starter or Growth for registered nonprofits and schools. Ask via /request-demo?nonprofit=yes with proof of status; approved within 1 business day.

What's white-label pricing for agencies?

Enterprise tier includes reseller pricing: you sell ChatGenius under your brand, we bill you at wholesale, you set your own margin. Contact us for a partner deck at partners@chatgenius.net.

Can I pay annually?

Yes — 12 months up front, 15% off across all tiers (and the BYO add-on). Invoiced by year, paid via bank transfer or card. Growth + Enterprise tiers can also be billed on custom cycles (quarterly, per-event).

What about taxes / VAT?

Prices are listed exclusive of applicable sales tax / VAT. EU customers see VAT added at checkout at their local rate; valid VAT ID zero-rates B2B invoices. US customers see any state/local sales tax at checkout. Enterprise contracts are net-of-tax by default.

Custom for chains, franchises, agencies.

Deploying ChatGenius across 20 hotel properties, 200 tour operator customers, or under your own brand? We do custom SLAs, regional hosting, dedicated support, white-label reseller agreements.