BETA · privacy LLMs & voice servers operational · GPU upgrade underway for faster responses · packages may still change Status & Roadmap →
GDPR-proof · EU-sovereign · No credit card to start

Save 40 hours per month on customer inquiries.Our AI answers 24/7 — from €9.95.

An AI helpbot that answers your customers instantly — from your knowledge base, in your tone of voice, in 6 main languages + 95 others. Saves the average SMB 40–130 hours of customer service per month, without hiring or firing anyone.

No credit card. Cancel monthly. First 100k tokens free with all features.

🇪🇺 EU-sovereign 🛡️ GDPR-proof + ISO 27001 🌍 24/7 multilingual helpbot ⚡ Live in 1 day

🛡️ Privacy by design — not an afterthought

ZelixAI is built for European businesses. Your customer data stays within the EU, in a dedicated isolated database for your tenant only.

🗄 Dedicated isolated database

No shared data pool. Every tenant gets a physically separate database — your customer data and conversations share no table with other customers.

📥 Full data export rights

Your data is your data. One click exports all your conversations, customer data, knowledge base and Q&A pairs as JSON or CSV. No vendor lock-in.

🇪🇺 EU-hosted

EU-hosted by default. No US cloud, no overseas data transfer. Our infrastructure runs on EU-sovereign servers (NL/FR datacenters).

📜 GDPR-compliant & ISO 27001

Data Processing Agreement (DPA) included by default from the first paid plan. For enterprise: BAA + ISO 27001 report + annual audit reports.

🔒 Privacy mode with one click

Switch with one click to a local LLM model — hosted on your own hardware or in our datacenter. 100% within your network perimeter.

🇳🇱 NL sovereign option

No US models wanted? Choose Mistral via OVH (NL-hosted) or a local LLM on your own hardware or in our data center. A fully EU chain.

Topic What you get From which plan
EU hostingDefault, no US cloudAll plans (incl. free)
Dedicated isolated tenant databaseNo shared data poolAll plans
Full data export rightJSON/CSV export of conversations, KB, customer dataAll plans
Data Processing Agreement (DPA)Standard signed PDFAll paid plans
Privacy kill-switch (local LLM)1-click switch, owner-locked, local LLM on your own hardware or in our DCAll plans (incl. free)
ISO 27001 reportAnnual audit reportSovereign + Custom
BAA / Business Associate AgreementFor healthcare, finance, legalCustom (€999 setup + €799/mo)
On-premise deploymentYour own servers, your own perimeterCustom (€999 setup + €799/mo)

💶 Calculate your own savings

Three concrete scenarios. Decision-makers don't calculate — they want 10 seconds to see: is this profitable?

A customer service rep averages ~€28/hour in labor costs (incl. employer charges). The average SMB spends 40-160 hours/month on customer inquiries. ZelixAI's AI helpbot typically deflects 60-70% of repeat questions — automatically.

Freelancer / small webshop

40 hrs customer inquiries/mo · €30/hr · 60% AI deflection

€720

saved per month · Chat Spark €9.95/mo · ROI 72×

SMB with support team

120 hrs customer inquiries/mo · €30/hr · 70% AI deflection

€2.520

saved per month · Business Launch €49/mo · ROI 51×

Scale-up / mid-market

320 hrs customer inquiries/mo · €35/hr · 70% AI deflection

€7.840

saved per month · Business Scale €249/mo · ROI 31×

Benefits not captured in the numbers

⚡ No 2-week course for your team

Your employees keep working as they are used to. The AI helpbot learns from their answers.

Your employees keep working as they are used to. The AI helpbot learns from their answers — no mandatory training, no external consultant, no "AI implementation project" taking 3 months.

How it works for your team

  1. Day 1: we automatically load your existing FAQ, product documents and website
  2. Days 2-7: AI answers 80%, your team checks the first 50 answers
  3. From week 2: AI handles 60-70% fully independently, team focuses on complex cases

No rocket science. No weeks of training. But an AI helpbot built into your portal that gives your team immediate answers in 6 main languages + 95 others.

What it is NOT

  • ❌ Not a replacement for your team — but relief from repetitive work
  • ❌ Not a "black box AI" — you see which question got which answer and can correct
  • ❌ No technical integration hassle — works out-of-the-box with WordPress, Shopify, Magento, custom sites
  • ❌ No 6-month contract — cancel monthly, first month free

💬 For you when… you mainly receive questions via your website

What you solve: customers who come to your site in the evenings or weekends and can't find an answer. AI responds directly from your knowledge base, creates tickets for what the bot doesn't know, keeps customer history.

Suitable from: Freelancer with 50 questions/month (Discover €0) to agency with 25 client bots (Agency €199).

📧 For you when… your customers also contact you via email and WhatsApp

What you solve: overflowing inboxes where your employee spends 4 hours/day answering emails that all look the same. AI sorts, answers or suggests replies — all under your control.

Suitable from: SMB with 1-3 customer contact channels (Launch €49) to enterprise with dedicated SLA (Sovereign).

ZelixAI Chat — widget-first helpbot for your website

Discover Chat

€0 — forever
300k tokens hard cap

"Try everything, pay only when you grow"

For whom: Freelancers and small websites wanting to test if AI can handle customer queries.

What it delivers: ~75 customer conversations/month answered automatically. Approximately 10-15 hours/month saved.

🤖 1 bot 💬 300k tokens 🌍 1 site

Real trial — all features enabled, no subscription fees

Included in this plan:

  • Full feature set, nothing disabled
  • 1 chatbot for your website (widget on one domain)
  • Website crawler: 1 page · Product crawler: 1 product
  • Q&A generator: AI automatically creates Q&A pairs
  • Own knowledge base: 1 document (~25KB text)
  • Create tickets + customer history
  • WordPress plugin for 1-click installation
  • 300,000 tokens/month (~75 conversations) — chat pauses after that
  • ZelixAI logo visible in widget until first payment

Chat Spark

€20/mo

"Cheaper than a Spotify subscription · saves half a workday per week"

For whom: Freelancers, small webshops, one-person businesses.

What it delivers: ~€720/month saved with 40 hours of customer queries/month · 60% AI deflection (€30/hour).

🤖 1 bot 💬 2M tokens 🌍 1 site

🎨 Custom branding — ZelixAI logo removed, your own colors

Included in this plan:

  • Everything from Discover, without ZelixAI branding in widget
  • 2M tokens/month (~500 customer conversations)
  • Website crawler: up to 50 pages · Product crawler: up to 100 products
  • Q&A generator: automatic Q&A pairs from documents + crawl results
  • 5 documents in knowledge base (~125KB text, approx. 50 A4 pages)
  • Own colors, own text — widget looks 100% your brand

Chat Studio

€139/mo

"Your clients together pay €69 — you pocket €350+/month margin"

For whom: Small marketing agencies, web agencies with 5-8 clients.

What it delivers: Charge €50/client/month — that is ~€330 margin on top of your time.

🤖 8 bots 👥 2 teams 💬 16M tokens

🏷️ Reseller portal — 1 login, 8 client dashboards

Included in this plan:

  • 8 independent bots — each client with their own widget, knowledge and branding
  • 2 teams for client segmentation
  • 16M tokens/month (~4.000 customer conversations total)
  • Website crawler: up to 2,000 pages per bot · Product crawler: up to 5,000 products per bot
  • Unlimited documents · up to 5MB text per bot (~2,000 A4 pages per bot)
  • Reseller portal: 1 login, 8 dashboards

Chat Pro

€239/mo

"15 clients at €7.93 per client · better margin than Crisp Reseller"

For whom: Agencies with 10-15 active clients in portfolio.

What it delivers: Charging €50/client/month: ~€630/month agency margin.

🤖 15 bots 👥 3 teams 💬 40M tokens

🏢 White-label — custom domain per bot (chat.klant1.nl)

Included in this plan:

  • 15 bots — agency sweet spot
  • 3 teams for industry segmentation
  • 40M tokens/month (~10.000 customer conversations total)
  • Website crawler: up to 5,000 pages per bot · Product crawler: up to 10,000 products per bot
  • Unlimited documents · up to 15MB text per bot (~6,000 A4 pages per bot)
  • Per-bot custom domain (chat.client1.com, chat.client2.com)
  • White-label options — clients see your agency, not ZelixAI

Chat Agency

€399/mo

"25 clients · €7.96 per client · top of the Chat line"

For whom: Larger agencies, marketing agencies, white-label resellers.

What it delivers: Charging €40/client: ~€800/month agency margin and SLA agreements for clients.

🤖 25 bots 👥 5 teams 💬 100M tokens

📞 Priority support + 99.5% SLA · unlimited crawl

Included in this plan:

  • 25 bots — full-portfolio agency tier
  • 5 teams
  • 100M tokens/month (~25.000 customer conversations total)
  • Website crawler: unlimited pages · Product crawler: unlimited products
  • Unlimited documents · up to 50MB text per bot
  • SLA targets: response
  • Priority support: WhatsApp + email (multilingual, business hours)

ZelixAI Business — unified inbox + helpbot for customer service teams

Discover Business

€0 — forever
300k tokens hard cap

"Try the full support platform — not just a demo, but a real trial"

For whom: SMB wanting to test if AI auto-reply and unified inbox really work.

What it delivers: ~75 emails or chats answered automatically — free up half a workday.

🤖 1 bot 👥 1 team 💬 300k tokens

Real trial — Unified Inbox + AI auto-reply fully enabled

Included in this plan:

  • Full feature set, nothing disabled
  • 1 widget bot + 1 unified inbox (Email/WhatsApp connected)
  • 1 team with 1 employee
  • AI auto-reply and inbox rules — 300k tokens (~75 conversations) to test
  • Website crawler + Product crawler + Q&A generator — 1 page / 1 product / 1 document
  • No surprises: at 300k service pauses until upgrade or credits

Business Launch

€69/mo

"Save your employee 4 hours/day on repetitive emails"

For whom: SMB with customer service via website + email + WhatsApp.

What it delivers: ~€2,520/month saved with 120 hours of customer queries/month · 70% AI deflection. ROI 51× on subscription.

🤖 2 bots 👥 1 team 💬 4M tokens

✉️ Email template wizard + inbox rules · AI replies or suggests

Included in this plan:

  • Unified Inbox: email + WhatsApp + chat widget in one
  • AI auto-reply: bot answers emails or suggests replies
  • Inbox rules: route emails by sender, subject, keywords
  • Mail template wizard — AI writes templates for you
  • 2 bots, 1 team, knowledge base sharing via team
  • 4M tokens/month (~1.000 emails or conversations)
  • Website crawler: up to 100 pages · Product crawler: up to 250 products
  • 10 documents in knowledge base (~250KB text)

Business Scale

€299/mo

"BYOK tier · pay AI costs directly at Anthropic or OpenAI · replaces 1-2 FTE on repetitive work"

For whom: Companies with 10-50 employees in customer success, with own AI budget and compliance requirements.

What it delivers: ~€7,840/month saved with 320 hours of customer queries/month. ROI 31×. Equivalent to not needing to hire 1 full-time employee.

🤖 12 bots 👥 4 teams 💬 30M tokens

🔑 BYOK per bot — no markup on tokens · agent ACL per team

Included in this plan:

  • BYOK per bot: connect own Anthropic, OpenAI, Gemini or OVH-Mistral key — no markup on tokens
  • 12 bots, 4 teams — full multi-department
  • Agent ACL per bot team (Sales data not visible to Support)
  • Audit log export
  • 📞 Callback appointments: calendar + reminders + magic-link cancel (max 200/mo)
  • SLA:
  • 30M tokens/month (~7.500 emails or conversations)
  • Website crawler: up to 5,000 pages per bot · Product crawler: up to 10,000 products per bot
  • Unlimited documents · up to 25MB text per bot

Business Sovereign

€499/mo

"EU-sovereign deal · BYOK across all providers · unlimited scalability"

For whom: Larger companies with strict compliance requirements wanting to scale without per-bot limits.

What it delivers: Unlimited scale headroom + signed DPA + EU-only routing — fits compliance budget.

🤖 ∞ bots 👥 ∞ teams 💬 100M tokens

🇪🇺 EU multi-region failover · white-label up to DNS · DPA included

Included in this plan:

  • Unlimited bots, teams, knowledge base · 100M tokens/month (~25,000 emails or conversations)
  • BYOK across all providers (Anthropic, OpenAI, Google, OVH, OpenRouter)
  • EU multi-region failover — no US cloud, no vendor lock-in
  • White-label to DNS level — own domain for portal and widget
  • 📞 Callback appointments: unlimited — agenda, reminders, magic-link cancel, iCal sync + webhooks
  • SLA:
  • DPA + ISO 27001 reporting standard
  • Priority email + WhatsApp support (business hours)
  • Website crawler + Product crawler: unlimited · up to 100MB text per bot

Business Custom

€999 one-time + €799/mo.

"Talk to us for your own package · full SLA · free support at integration"

For whom: government, healthcare, legal, finance, large retailers — situations where standard SLAs are not sufficient.

What you get: personal meeting with our experts, tailored quote, and ZelixAI fitting your IT architecture.

Included in this plan:

  • €999 one-time — integration work, knowledge base import, API connection with your systems, training for your employees
  • €799/month — unlimited scale, all features from Sovereign + the extras below
  • Full SLA: response
  • BAA + annual audit reporting + on-premise deployment
  • Free support at integration — our engineers help your IT team
  • Dedicated account manager — direct via chat, email, video call
  • 24/7 monitoring + proactive incident notification
  • On-premise deployment — run ZelixAI on your own servers
  • Custom contracts — volume pricing, multi-year deals, custom DPA clauses: everything negotiable
Schedule a meeting →

Response within 1 business day. "Custom is not a subscription — it's a service relationship."

📞

Customers book their own callback appointments — via chat or email

The chatbot or email bot recognises when a customer wants a callback. It asks for their name, phone number and reason, shows available time slots and books the appointment directly. Your team sees all appointments in a dedicated agenda — including drag-and-drop rescheduling, automatic reminders and a magic link the customer can use to cancel.

  • Included from Business Grow (50/mo) · Scale (200/mo) · Sovereign (unlimited)
  • Add-on for Chat Boost & Business Launch: €14.95/mo (25 appointments/mo)
  • Confirmation and reminder emails · iCal sync with Google / Outlook / Apple Calendar · Webhook for Zapier and your own CRM

🔧 Technical details (for those who want to go deeper)

Not relevant for the decision — but fully transparent. Click to view.

🤖 Which AI models are available? (400+ models)

ZelixAI is a meta-platform — you choose per bot which AI model to use. Available providers:

  • Anthropic Claude — Opus 4.7, Sonnet 4.6, Haiku 4.5 (best for tone of voice)
  • OpenAI GPT — GPT-5, GPT-4.1, GPT-4o (industry standard)
  • Google Gemini — Pro, Flash (best in cost-per-token)
  • ★ ZelixAI Best Choice Mix — Mistral via OVH — Large, Small (fully EU-sovereign, NL-hosted) — our recommended default for EU compliance
  • ★ ZelixAI Best Choice Mix — OpenRouter — 369 alternative models (Llama, DeepSeek, Qwen, Mistral-open) — our recommended routing for the widest choice + best price-performance
  • Local LLM model — hosted on your own hardware or at our data center, for privacy mode and strict data residency

Our "Best Choice Mix" combines OVH (EU-sovereign) with OpenRouter (widest choice) — giving you automatically the right model per bot for the right task. Switch manually or let ZelixAI's router handle it — in 1 click.

Best-Choice Routing: ZelixAI advises per use case

ZelixAI analyses your use case during bot configuration and recommends which model best fits. You are free to choose — the advice is indicative.

Use case Recommended model Reason
NL customer service (simple questions) Mistral via OVH EU-sovereign, NL-hosted, excellent NL comprehension
Multilingual customer service GPT-4o-mini Strongest multilingual performance, low cost
Complex Q&A / legal / technical Claude Sonnet Best reasoning and tone-of-voice quality
High-volume simple FAQ Claude Haiku / Gemini Flash Lowest cost per token, fast
Privacy-first / no external calls Local LLM Own hardware or our data center, no external API

Model evaluation framework

Every model in our stack is continuously benchmarked on 5 dimensions: (1) language comprehension NL/EN/DE/FR, (2) tone-of-voice preservation, (3) factual accuracy on knowledge base content, (4) refusal rate on out-of-scope questions, (5) latency p95. Results drive our Best-Choice-Routing recommendations.

🔑 How does BYOK (Bring Your Own Key) work?

From Business Scale (€249/month) you can connect your own API key per bot for Anthropic, OpenAI or other providers. You then pay tokens directly to the provider — we only charge the platform fee.

Benefits: lower AI costs if you have an Enterprise contract at Anthropic/OpenAI · one source of API spending for your accountant · no ZelixAI markup on tokens.

👥 What are bot roles and bot teams?

A bot specialises in one topic (Sales, Support, Tech, Billing). Bot teams group multiple bots so they can:

  • Share a common knowledge base (all Sales bots see each other's products)
  • Receive an employee access layer (ACL) — Support team only sees Support data
  • Escalate between each other — Sales bot can automatically hand over a tech question to Tech bot

Compliance-by-design for healthcare, legal, finance: a Sales bot cannot structurally see patient data if it sits in the Support team.

🛡️ How does privacy mode work (3 levels)?

ZelixAI has three privacy levels — owner-locked, only the tenant owner can switch, audit log records every change.

LevelWhat it doesSpeed / quality
Standard All providers allowed. EU routing bias where possible. Good default for most SMBs. Full — all 400+ models
Strict Mode Only models with signed DPA + no-training clause + EU data residency: Mistral via OVH (NL), Google Vertex AI EU, AWS Bedrock EU, local LLMs. Cloud speed, cloud quality, with privacy assurance
Local-only No external LLM calls at all. Only local LLM model hosted on your own hardware or at our data center. Absolute certainty. Slightly slower, smaller models — 100% guaranteed
📦 How does token billing and the bundle work?

Every chat consumes tokens (approximately 1 token = 0.75 words). Your plan includes a bundle (e.g. Chat Spark = 2M tokens). You receive notifications at 80% and 100% usage.

  • Under the bundle: everything flat — no surprises, you pay your fixed monthly price
  • Scale discount on growth: the larger your plan, the lower the per-token price on peaks. Growth automatically becomes cheaper.
  • Discover (free): hard cap at 300k tokens (~75 conversations) — chat pauses until you buy credits or upgrade. No surprises.
  • Opt-in pre-paid credits: available on all paid tiers (Spark through Custom). Activate in dashboard, pay upfront via iDEAL/CC/SEPA. No extra tiers required.

Exception: Opus models

Anthropic Claude Opus is the most powerful (and most expensive) model in our stack. For Opus, a flat 50% markup reduction across all tiers applies — regardless of plan. So: Discover users who select Opus pay the same percentage surcharge as Scale users on Opus. This is intentional: we want you to choose the best model based on the use case, not based on tier. For most customer service tasks, Haiku-equivalent already delivers excellent results; Opus makes sense for complex legal or technical Q&A.

🔌 How do I integrate ZelixAI with my existing website?
  1. WordPress plugin — download via your dashboard, install in WP admin, ready in 5 minutes
  2. JavaScript widget tag — copy 1 line of JS into your , works on any site (Shopify, Magento, custom)
  3. API integration — REST API for backend integrations or custom frontend

Custom widget domain from Chat Studio (€69/month) — so the chat becomes chat.yoursite.com instead of widget.zelixai.ai.

Why ZelixAI?

Not because we say so — but because you can measure it yourself.

⏱ Saves half an FTE

The average SMB saves 80-130 hours of customer service per month with our AI. Equivalent to half a full-time employee — without hiring, without firing, without compromising on service.

🇪🇺 Fully EU-sovereign

Your customer data stays in the Netherlands, in a dedicated isolated database for your tenant only. GDPR-proof by design, ISO 27001 tested. One click switches to local LLM. Full export right on your data — no vendor lock-in.

🌍 Always-on multilingual support

Our AI helpbot answers in 6 languages — Dutch, English, German, French, Turkish, Spanish — free, in your language, 24/7. Email support also in your language.

Frequently asked questions

Can't find your question? Contact us — we respond within 1 business day.

In our beta the AI answers 60-70% of customer questions fully autonomously when your knowledge base is well set up. The remaining 30-40% the bot forwards as a ticket or suggests an answer your employee confirms with one click. The percentage rises to 80%+ after 2-3 months — the bot learns from corrections.

No. Your existing FAQ, product documents and website are automatically loaded on day 1. Your team keeps working as always — just with fewer repeat questions. An AI helpbot built into your portal explains directly to employees in 6 main languages + 95 others how it works.

Standard EU-hosted (Netherlands), in a dedicated isolated database per tenant. Data Processing Agreement (DPA) included from the first paid plan. Privacy kill switch on all plans — disables cloud LLM, bots then run on a local LLM model hosted on your own hardware or at our data center. For healthcare/finance/legal: BAA + audit reporting + on-premise deployment from Business Custom.

Three built-in safeguards: (1) AI only answers from your knowledge base — no "made-up" facts; (2) when in doubt the bot forwards the question as a ticket instead of guessing; (3) you see all answers in your dashboard and can correct — the bot learns from it.

For WordPress sites: 5 minutes (install plugin, done). For other sites: 15 minutes (1 line of JS in your ). Knowledge base is automatically populated from your existing site, FAQ and documents — first answers within the hour of going live.

On paid plans everything continues normally. We warn at 80% and 100% of your bundle. Above that a scale discount rate applies that decreases as your plan gets larger — growth automatically becomes cheaper. Pre-paid credits are opt-in available on all paid tiers (Spark through Custom) — no penalty, no interruption. On Discover (free) the service pauses at 300k tokens (~75 conversations) until you buy credits or upgrade.

Dutch tenants: iDEAL, SEPA direct debit and credit card via Pay.nl — automatic monthly billing. EU tenants outside NL: SEPA direct debit, credit card and PayPal via Stripe. Tenants outside EU: credit card, PayPal and wire transfer. All payments are monthly and pro-rated for upgrades.

Cancel monthly, no minimum contract (except Business Custom: annual contract). Pro-rated billing for the current month on upgrade or downgrade. Switching between Chat and Business is possible at any time — you keep all your bots, knowledge base and history. No migration, no downtime.

Yes — simple and complete. One click in the admin panel exports all your conversations, customer data, knowledge base articles and Q&A pairs as JSON or CSV. No exit fees, no "contact us" hassle, no vendor lock-in. You pay for service, not for imprisonment.

Try it for free. The Discover tier stays permanently free with all features — you can test for a month whether 60-70% of your repeat questions get answered automatically. Only then do you upgrade. No credit card to start.

Save the first 8 hours this month already.

Start free with all features. First 300k tokens (~75 conversations), no credit card, no migration pain. Upgrade only when you grow — no downtime, no hassle.

Start free with ZelixAI Chat → Start free with ZelixAI Business → Or schedule a demo (15 min)

Questions? Email support@zelixai.ai or ask our AI helpbot — answer in your language within seconds.