Pricing
Plans that scale with you
Start with a 3-month pilot
Same features, full access. Cancel with 7 days' notice.
Start with a 3-month pilot. Same features, full access, cancel with 7 days' notice.
Pricing
Plans that scalewith you.
Starter
$449/mo
For new customers signing up in 2026
2,000 conversations included / mo
- 1 AI Agent
- Certainly AI included
- Web widget channel
- Connect your knowledge base and website
- Integration builder
- CRM integration
- Unlimited seats
- Traffic & Business Reporting
- Video onboarding
- Community & help centre
- Email support
Pro
$1,099/mo
For new customers signing up in 2026
10,000 conversations included / mo
Everything in Starter, plus:
- 2 AI Agents
- Bring your own key
- 1 additional channel (FB, IG, WhatsApp or Email)
- Personalised CRM integrations
- AI reporting (up to 3 reprocesses / month)
- Dedicated onboarding (€800 value)
- Pooled TAM support
Enterprise
Bespoke pricing
Tailored to your requirements
30,000+ conversations included / mo
Everything in Pro, plus:
- Unlimited AI Agents
- Reporting & Chat API access
- All channels supported
- White-label options
- Custom SLA with uptime credits
- Dedicated account manager
- AI reporting (up to 10 reprocesses / month)
- Enterprise onboarding (€2,000 value)
- Priority engineering support
Need someone to handle the technical side? We know your internal teams are busy. Our implementation specialists build, integrate, and optimise your AI agent end‑to‑end so you get results without waiting on dev resources.
Talk to our teamHow many conversations do you handle each month?
Drag the slider for a rough plan and savings estimate. Real pricing depends on channels, integrations, and AI usage.
Best fit
Pro
Est. Certainly cost
€1,250/mo
Est. monthly savings
€4,901(82%)
Savings vs. fully human-handled tickets at €1.20 average handling cost. AI usage billed at provider rates plus 20% surcharge.
“Certainly did not give us a chatbot. They gave us a revenue channel. The agent sounds like us, sells like us, and our customers genuinely enjoy talking to it.”
Jess Cervellon
VP of CX, Feastables
- 4.8 on G2 Read reviews
- Capterra reviewed
- Gartner Peer Insights
- SOC 2 aligned
Where we sit
The space between simple and bespoke.
CRM chatbot
Simple. But limited.
Plug-and-play, quick to deploy. But it follows a script. It cannot reason, act, or connect deeply with your systems. When your customer goes off-script, it breaks.
Certainly.
The complete platform.
- ✓Agentic AI that acts
- ✓Every channel, connected
- ✓Your stack, intact
- ✓Live in days, not months
- ✓Supported from day one
LLM / bespoke build
Powerful. But unfinished.
Highly capable AI, but no platform around it. No channels. No integrations. No support. You own every connection, every deployment, every problem.
Side-by-side
Evaluating Certainly against other AI agent platforms?
Honest, line-by-line breakdowns of how Certainly compares to the platforms you are likely already evaluating.
vs Intercom Fin
LLM agnostic vs single-model lock-in
Read comparisonvs Ada
Transparent pricing vs custom quotes
Read comparisonvs Zendesk AI
Automation specialist vs CRM add-on
Read comparisonNot sure which plan?
Answer 4 quick questions and we'll recommend the best fit.
How many conversations does your team handle monthly?
Include all customer support interactions across channels.
How we price, and why
Per conversation, per resolution, or per token?
The pricing model shapes the incentives more than the headline rate does. Here is the honest comparison so you can defend your choice to the CFO in one sentence.
Per conversation
A flat unit per handled interaction. Predictable from your ticket volume, with no debate about whether a refund, escalation, or clarification counted.
- + Forecastable from existing ticket data
- + Same unit finance and operations both see
- + Token-level model cost visible in the dashboard
Per resolution
You only pay when the AI reaches a defined outcome. Aligns to value when the definition is transparent, but the definition usually sits with the vendor.
- − Resolution definition is vendor-controlled
- − Edge cases, reopens, and escalations get billed unevenly
- − Hard to reconcile against your own systems
Per token
What model providers actually charge for. A token is roughly three quarters of a word. The most auditable unit, but it varies with prompt length, model, and reasoning depth.
- + Most auditable line item, matches provider invoices
- − Hardest to forecast without model routing discipline
- = Surfaced inside every Certainly conversation
The vendors who cannot explain the unit they sell are pricing for their margin, not your economics. The three questions to ask any AI customer service vendor: who defines the billable unit, what happens to escalations and reopens, and can the account team explain the underlying token cost.
Read the full breakdownPricing FAQ
Clear answers on conversations, reprocessing, and what is included.