Private Office
Your everyday business tools on one private server. Add a compatible provider API key for AI answers; without one, the rest of Private Office still works.
Pricing
One flat price. Your whole business.
All your day-to-day tools in one place: files, documents, customers and invoicing, on a European server reserved for your organisation. One fixed price, not a separate bill for every app. Private Office starts at €39/mo; Managed EU AI will soon be available as an optional add-on from +€19/mo.
Plans
Start with Private Office. Add Managed EU AI when you need it.
Working on your own? Choose Solo. Have a team? Team brings up to 10 people into one shared workspace at a fixed price. Managed EU AI will be an option if you do not want to create or manage an API key yourself.
Founding offer — lock in the founding price before the end of 2026, for life: €39/mo Solo · €99/mo Team
Save even more on annual plans — €29/mo Solo · €79/mo Team (billed annually · save €120/€240 per year)
Private Office Team at €99/month — €9.90 per person for a 10-person team. No per-seat fees on any plan.
Recommended first team plan Most popular
Private Office Team
€79 €99 /mo
€948 billed once a year Save €240 per year
Founding price excl. tax — locked for life
Best first plan for small teams — up to 10 named users, flat rate
Under the hood ▸
Contabo Cloud VPS 8 · 8 vCPU AMD EPYC · 24 GB RAM · 300 GB SSD · EU datacenter
- ✓Files, passwords, documents & automation — all included, nothing to configure
- ✓300 GB included · paid 600 GB option offered in the next step, before payment
- ✓+ Centralised member onboarding, MFA and offboarding
- ✓Up to 10 named users · flat rate · 5 TB backup · daily security scans
- ✓Starter automation recipes included: client intake folders, document summaries, deadline reminders, weekly reports
- →The system and apps use part of the disk. Need more later? Add 5 TB encrypted expansion storage on request; the dashboard shows free space and warns you early
Included AI tools — with a compatible API key
- ✓Open WebUI: summarise invoices, review contracts and find key dates.
- ✓Hermes: prepares drafts and guides you through tasks; actions stay under your control.
- →What is a compatible API key?
- →No API key to manage? Discover Managed EU AI.
Managed EU AI
Managed EU AI
AvailableAdd fully managed AI hosted in Germany to an active Private Office — 30-day Essential trial, no GPU or API key to manage.
Essential· 40 M tokens/mo +€39/mo
Chat, document work and governed Hermes actions; writes require approval by default
Plus· 120 M tokens/mo +€99/mo
The same safe capabilities with a larger monthly token allowance
Expert· 180 M tokens/mo+40 M high-capacity/mo +€169/mo
Frontier escalation model with its own allowance · governed tools and persistent learning
monthly add-on · billed separately from your Office plan
Fair-use token allowance with a hard cap — no surprise overage billing. Unused allowance rolls into the next month.
What does this mean in practice?
With 40M tokens: answer about 2000 client emails, search about 400 documents, or run about 200 automated agent tasks per month. IONOS AI Model Hub currently processes prompts in Germany.
What the add-on adds
- ✓Chat, document Q&A and retrieval across your files — files and index stay on your server.
- ✓A stronger cloud model for Hermes and your automations when the built-in local AI isn't enough.
- ℹ Prompts and selected excerpts go through an encrypted tunnel to IONOS AI Model Hub in Germany; your files, indexes and history stay on your server.
- Your Office already includes local AI. This adds managed EU cloud AI when you want more headroom — or connect your own model key instead.
- →Need guaranteed capacity? Join the Dedicated AI validation programme →
Dedicated AI
PilotSupplier and capacity validation pending
Managed EU AI is a waitlisted option that will use contractually approved European routing without you managing hardware. Private Dedicated AI is a separate assisted offer — a dedicated GPU sized and quoted for your workload and isolation needs.
- ✓Hermes AI agent included by default on your dedicated setup
- ✓Your GPU — no other client's workload, ever
- ✓For legal, healthcare, finance — AI processing that never touches shared hardware
- ℹSupplier and capacity validation pending
Brings together functions often bought separately. Your real saving depends on the tools you replace: compare ReefOffice with your current basket, not a generic estimate.
Compare plans
What changes between plans
Every plan gives you your own dedicated server in the EU — your machine, your storage, your data, isolated from all other customers.
| Feature | Private Office Team €79/mo Founding price excl. tax — locked for life Teams up to 10 — collaboration, flat rate, local AI Apply for a pilot → | Private Office + Managed AI From +€39/mo Add EU-hosted AI to your Office Start with Private Office | Dedicated AI Supplier and capacity validation pending Supplier and capacity validation pending Join the waitlist → |
|---|---|---|---|
| 🗄️Workspace | |||
| EU dedicated server | ✓ | ✓ | ✓ |
| Files, calendar, contacts & editing | ✓ | ✓ | ✓ |
| CRM & invoicing | ✓ | ✓ | ✓ |
| Password manager | ✓ | ✓ | ✓ |
| Document archive (OCR) | ✓ | ✓ | ✓ |
| No-code automations | ✓ | ✓ | ✓ |
| Encrypted backup | 5 TB | 5 TB | Supplier and capacity validation pending |
| 👥Team & Collaboration | |||
| Encrypted messaging + video | ✓ | ✓ | ✓ |
| Shared folders & permissions | ✓ | ✓ | ✓ |
| Users | 10 | 10 | Supplier and capacity validation pending |
| 🤖AI & Search | |||
| Smart search in your docs | ✓ | ✓ | ✓ |
| Managed AI — Essential | API key from your chosen provider (Mistral, OpenAI, Anthropic, etc.) | +€39/mo Chat: Mistral Small 24B Agent: Mistral Small 24B 40M tokens/mo Chat, documents and governed Hermes actions; writes require approval by default | Dedicated AI |
| Managed AI — Plus | API key from your chosen provider (Mistral, OpenAI, Anthropic, etc.) | +€99/mo Chat: Mistral Small 24B Agent: Mistral Small 24B 120M tokens/mo Same capabilities with more monthly tokens | Dedicated AI |
| Managed AI — Expert | API key from your chosen provider (Mistral, OpenAI, Anthropic, etc.) | +€169/mo Chat: Mistral Small 24B Agent: Mistral Small 24B Advanced tasks: gpt-oss-120b 180M tokens/mo + 40M high-capacity/mo Frontier escalation model with its own allowance · governed tools and persistent learning | Dedicated AI |
| ⚙️Infrastructure | |||
| Processor | 8 vCPU | 8 vCPU | Supplier and capacity validation pending |
| RAM | 24 GB | 24 GB | Supplier and capacity validation pending |
| SSD Storage | 300 GB included 600 GB paid option at setup | 300 GB included 600 GB paid option at setup | Supplier and capacity validation pending |
| Graphics processor (GPU) | — | — | Supplier and capacity validation pending |
| 🔒Security & Trust | |||
| 30-day money-back guarantee | ✓ | — | — |
| Open source — you own everything | ✓ | ✓ | ✓ |
| No exit fees | ✓ | ✓ | ✓ |
| GDPR by default | ✓ | ✓ | ✓ |
| AI processing — ISO 27001 datacenter (Germany) | — | — | Supplier and capacity validation pending |
| 🛟Support | |||
| Technical support | Reply within 24h | Reply within 24h | Reply within 24h |
| Data export | ✓ | ✓ | ✓ |
| SSO (Authentik) | ✓ | ✓ | ✓ |
Every plan includes
- A dedicated EU server, reserved for you — never shared.
- Files, calendar, contacts, documents, CRM & invoicing.
- Daily encrypted backups in the EU.
- Hermes agent included — use a compatible API key, or join the Managed EU AI waitlist.
- Open-source stack, full export, no lock-in.
- Storage that grows with you — encrypted expansion available on every plan.
Managed EU AI
A waitlisted add-on to any Private Office plan: managed chat and agent models through a contractually approved European provider.
Dedicated AI
Managed EU AI is a waitlisted option that will use contractually approved European routing without you managing hardware. Private Dedicated AI is a separate assisted offer — a dedicated GPU sized and quoted for your workload and isolation needs.
Hosted in the EU GDPR by default 30-day money-back guarantee No lock-in — export anytime
More than 10 users?
Teams of 11 or more need a capacity review so we can recommend the right server size. Contact us and we'll prepare a tailored quote — same flat-rate pricing, no per-seat fees.
Contact us for larger teams →Payment
We accept SEPA direct debit and card payments (Visa, Mastercard, Cartes Bancaires), processed securely by Mollie (Netherlands). French VAT is currently not charged under art. 293 B of the CGI. Annual billing saves ~20%.
Managed AI is available
Already have an active Private Office? Sign in to Billing to start a 30-day Essential trial. New to ReefOffice? Choose your Office first; Managed AI stays separate and optional.
Early feedback
"I could build my own workflow inside ReefOffice without technical knowledge, through Grist."
"Perfect for automating my workflows without needing heavy AI."
From early testers, quoted with their permission.
Learn more
Pricing questions
How many users are included?
Solo includes one named user. Team includes up to 10 named users at a flat price. Eleven or more users require a capacity review and assisted quote so the VM can be sized safely. Restricted Grist and Nextcloud guests do not consume a full team seat.
Monthly or annual — what's the difference?
Both are available on every plan. Annual is billed yearly and works out to a lower monthly rate. You can start monthly and switch later.
Is VAT added on top?
French VAT is currently not charged because ReefOffice operates under the franchise-en-base régime — TVA non applicable, art. 293 B du CGI. The checkout and invoice show the total amount payable; contact us for cross-border invoicing questions.
What if I want to leave?
You export everything with one click and can self-host the exact same open-source stack. There's no proprietary format and no lock-in.
What do 'Managed EU AI' and 'Dedicated' mean?
Managed EU AI is a waitlisted option that will use contractually approved European routing without you managing hardware. Private Dedicated AI is a separate assisted offer — a dedicated GPU sized and quoted for your workload and isolation needs.
Can I try it or see it first?
Yes. Try the isolated self-service demo free for two hours, or request a guided demonstration if you prefer to discuss your workflow with us.
What if I run out of space?
Every server ships with 300 GB — years of documents for most businesses — and your dashboard shows disk usage and warns you well before it fills. If you do outgrow it, encrypted expansion storage (5 TB and beyond) can be attached to any plan, Solo included. It's encrypted on your server before it leaves, so the storage provider only ever sees ciphertext — the same guarantee as your backups.
What happens if I use up my Managed AI token allowance before the end of the month?
Managed AI runs through a fail-closed monthly cap, so you can never be surprised by an overage bill. A typical solo user consumes roughly 0.5–1.5M tokens a month, while the Essential Solo allowance is 10M — several times a realistic month's load. If you do approach the cap, the dashboard shows your usage and unused allowance carries over into the next month (Essential rolls over up to 2× its allowance), so a heavy month is cushioned rather than cut off. You can upgrade tiers or add your own model key at any time; there is no automatic metered spend beyond your flat plan.
What is a compatible API key?
Private Office accepts a compatible API key. You create it separately with the AI service you choose: a regular ChatGPT, Claude or Le Chat subscription does not replace API access. Its price and handling of submitted text follow that service's terms. Prefer not to configure or monitor a key? Join the Managed EU AI waitlist and ReefOffice will manage the AI connection for you.
Ready to own your back-office?
Configure your Private Office Solo in minutes — card or SEPA, cancel anytime.