🇪🇺 EU infrastructure by default

Every Agent Needs a Box.Managed OpenClaw in Europe for teams

Deploy managed OpenClaw and hosted OpenClaw in minutes on EU infrastructure. Connect your tools via OAuth, add governance controls with audit logs, and keep inference in‑region with EU‑only mode.

Founding cohort · Rolling invites.

Copy a short note for your CTO or security lead.

🇪🇺 EU data residency (DE/FI)
GDPR‑aligned controls
NIS2‑aligned controls
SOC 2‑ready controls
Zero Data Retention option
EU AI Act‑aware

Copy a short note for your CTO or security lead.

Why Tidebox

Managed Agents-as-a-Service. Not a VPS tutorial.

Hosting OpenClaw is easy. Running it safely with tool access, policies, logs, and compliance evidence is the hard part. Tidebox gives you the controls enterprises ask for — without building a DevOps team.

Governance‑first

Audit logs, role‑based access, and configurable retention.

🇪🇺 EU‑first

Choose EU regions and keep data processing in‑region where required.

Agent supply‑chain safety

Curated skills, allowlists, and scanning.

Channels

Your agent goes where your team already works

No new apps to install, no new tabs to open. Tidebox agents live inside the tools your team uses every day — so adoption is instant and friction is zero.

💬Slack

Runs as a native Slack bot. Mention it in any channel, DM it, or let it post proactively.

📱WhatsApp

Business API integration. Your agent responds to customers and team members on WhatsApp with full audit trail.

✈️Telegram

Native bot integration. Deploy agents to Telegram groups or direct chats.

🟣Microsoft Teams

Full Teams app. Works in channels, group chats, and 1:1 conversations.

✉️Email

Gmail and Outlook integration. Your agent reads, drafts, and sends emails with scoped permissions.

📞Voice

EU phone number included in Teams plan. Your agent answers calls and speaks in your team’s language.

🔌API / Webhooks

Embed your agent in any internal tool, dashboard, or custom workflow via REST API.

🌐Web widget

Drop a chat widget on your site or internal portal. Zero‑code setup.

All channels share the same agent brain, permissions, and audit trail. One config, every touchpoint.
Vetted Skills & Connectors

Curated and vetted skills — in minutes, not months

Give your agents trusted capabilities without turning every integration into a security review. Install vetted skills, connect the tools you already use, and keep permissions explicit from day one.

Curated marketplace skills

Install curated and vetted skills from Orzogonal and other verified marketplaces. Connect your agent to Slack, GitHub, Jira, HubSpot, Zendesk, Linear, Gmail, Notion, and dozens more — with a single click.

Custom skills

Build your own skills in TypeScript or Python. Define inputs, outputs, and permissions. Deploy them to your workspace with full version control and rollback.

Vetting + security scanning

Every skill — marketplace or custom — is vetted before installation. We check known vulnerabilities, excessive permissions, and data leakage patterns. No ‘install anything and hope’ approach.

Scoped permissions

Each skill declares what it needs (read, write, API access). You approve explicitly. An agent with a Slack skill can’t silently access your GitHub repos.

All skills run inside the same isolated EU runtime as your agent. No data leaves your workspace boundary unless the skill’s declared scope requires it.
Prebuilt agents

Prebuilt agents for enterprise workflows

Start from enterprise-ready templates with prewired connectors already in place. Then tailor permissions, policies, and escalation rules for your team.

Incident Responder

Prewired connectors: Sentry + GitHub + Slack

Detects spikes, correlates deploys, opens incident threads, drafts rollback notes, and escalates with full context.

Security Questionnaire Helper

Prewired connectors: Notion + Docs + Slack

Maps buyer questionnaires to your controls, drafts answers, and routes exceptions for review.

Customer Support Triage

Prewired connectors: Zendesk + Slack

Classifies tickets, drafts replies, tags root causes, and escalates edge cases with context.

Linear Ticket Triage

Prewired connectors: Linear + Slack

Deduplicates issues, routes ownership, asks for missing details, and keeps product triage moving.

Deal Desk Assistant

Prewired connectors: HubSpot + Gmail + Slack

Prepares account context, drafts follow-ups, and coordinates internal approvals for enterprise deals.

Ops Digest

Prewired connectors: Gmail + HubSpot + Slack

Delivers a daily ops brief with anomalies, risks, customer signals, and recommended actions.

🇪🇺 EU-only inference

Keep inference inside Europe — without giving up frontier models

Some teams can’t send prompts outside the EU. Tidebox offers an EU‑only inference lane that routes requests to providers and regions configured for in‑region processing.

OpenRouter EU

EU base URL routing when enabled.

OpenAI Europe

EU project routing for eligible accounts.

AWS Bedrock EU

Claude via EU‑bound inference profiles.

EU‑only availability depends on provider, model, and your account eligibility.
Scope detail: What’s in‑EU: prompt routing, model inference, and runtime data. What may leave the EU: provider‑side telemetry and billing metadata (varies by provider). We document each provider’s guarantees in your workspace policy dashboard.
Security & governance

Security‑first, by design

Isolation

  • Dedicated runtime per workspace
  • Encrypted storage + in transit
  • Controlled updates (no ‘pull latest’)
  • Backups + rollback snapshots

Governance

  • RBAC: admin / operator / auditor
  • SSO (OIDC/SAML) + SCIM (Enterprise)
  • Audit log export (SIEM‑ready)
  • Retention controls (3 days → 1 year)
Read security details
Trusted by teams building with AI agents

We went from ‘one dev managing a VPS’ to production-grade agent governance in two days. The audit trail alone sold our compliance team.

Head of Engineering, Series B SaaS (DACH)

EU‑only inference was the dealbreaker for our legal team. Tidebox made it a toggle, not a six-month infra project.

CTO, RegTech startup (Nordics)
Design partner (stealth, DACH)
Design partner (stealth, Nordics)
Launching Q2 2026
Average time to first production agent:< 2 hours
Pricing

Pricing

Proposed launch packaging for the teams we are qualifying now. Each plan bundles runtime, governance, and inference so we can validate real willingness to pay before launch.

What’s inside each plan

Workspace
An isolated EU runtime environment with its own storage, secrets, and audit trail. Builder includes 1 workspace; Teams gives you 3. Each workspace is a secure sandbox for your agents.
Agent
A configured AI assistant running inside a workspace. Each agent has its own template, skills, permissions, and escalation rules. Teams plan: up to 3 agents per workspace (9 total across 3 workspaces).
Team member
A human who can build, configure, and manage agents via the dashboard. End‑users who interact with agents via Slack, Teams, voice, or other channels are always unlimited and free.
Skills
Pre‑built or custom capabilities your agents can use (Slack, GitHub, Jira, etc.). All skills enabled for your plan are available to all agents in the workspace.

Builder

For solo builders shipping their first agent.
€99/ mo
Billed monthly
€25 credits ≈ 400 calls/mo on Claude Sonnet or ≈ 60 on Opus 4.6
Managed inference included€25/mo
OveragePay-as-you-go
BYOK (own API keys)Supported
EU‑only inference laneIncluded
Spend hard capConfigurable
  • 1 workspace, 1 agent
  • EU runtime (DE or FI)
  • Unlimited end‑users (Slack, Teams, etc.)
  • All integrations & channels
  • All skills enabled
  • Templates + heartbeat monitoring
  • Skill allowlist + scanning
  • 7‑day audit log retention
  • Self‑serve docs + community support
Founding cohort · Rolling invites · No card required
Best for teams running 2‑3 production agents.

Teams

€299/ mo
Billed monthly
€95 credits ≈ 1,500 calls/mo on Claude Sonnet or ≈ 230 on Opus 4.6
Managed inference included€95/mo
OveragePay-as-you-go
BYOK (own API keys)Supported
EU‑only inference laneIncluded
Spend hard capConfigurable
  • Everything in Builder
  • 3 workspaces, up to 3 agents each
  • Up to 10 team members
  • Unlimited end‑users (Slack, Teams, etc.)
  • 30‑day audit log retention
  • Voice agents with EU phone number(beta)
  • Priority support + guided onboarding
Founding cohort · Rolling invites · No card required

Enterprise

For regulated teams needing full control.
Custom
Custom credit commit — we size it to your workload
Managed inference includedCustom commit
OverageNegotiated
BYOK (own API keys)Supported
EU‑only inference laneIncluded
Spend hard capConfigurable
  • Unlimited agents & workspaces
  • SSO/SAML + SCIM
  • 1‑year log retention
  • Dedicated infrastructure options
  • Custom workspace subdomains on tide.box
  • SLA + custom integrations
  • NIS2‑aligned incident pack
  • Dedicated account manager
Founding cohort · Rolling invites · No card required

Still undecided? Talk to us.

We’ll help you pick the right plan for your team, compliance needs, and workload.

How inference billing works

1. Included credits

Every plan ships with managed inference credits (Builder €25/mo, Teams €95/mo, Enterprise custom). Credits cover LLM calls routed through Tidebox.

2. Overage: pay‑as‑you‑go

When credits run out, usage continues at pay‑as‑you‑go rates. No surprise bills — you set a hard cap or auto‑stop per workspace.

3. Hard cap / auto‑stop

Configure a monthly spend limit. When reached, agents pause or fall back to BYOK. You stay in control.

Prefer full control? Use BYOK mode on any plan: plug your own OpenAI / Anthropic / OpenRouter API keys and pay providers directly. Managed credits and BYOK can coexist in the same workspace.

Estimate your inference

20
1
Estimated monthly inference
~€36
Covered by your plan: €25Extra pay‑as‑you‑go: €11

Indicative estimate only, not a quote. Actual costs depend on prompt length, output tokens, and provider pricing.

Add‑ons

Managed inference credit top‑up€10 per €10 block
Dedicated region / serverfrom €199/mo
Break-glass admin access (audited, no public SSH)Enterprise add-on
What’s inside each plan
Workspace
An isolated EU runtime environment with its own storage, secrets, and audit trail. Builder includes 1 workspace; Teams gives you 3. Each workspace is a secure sandbox for your agents.
Agent
A configured AI assistant running inside a workspace. Each agent has its own template, skills, permissions, and escalation rules. Teams plan: up to 3 agents per workspace (9 total across 3 workspaces).
Team member
A human who can build, configure, and manage agents via the dashboard. End‑users who interact with agents via Slack, Teams, voice, or other channels are always unlimited and free.
Skills
Pre‑built or custom capabilities your agents can use (Slack, GitHub, Jira, etc.). All skills enabled for your plan are available to all agents in the workspace.
How inference billing works

1. Included credits

Every plan ships with managed inference credits (Builder €25/mo, Teams €95/mo, Enterprise custom). Credits cover LLM calls routed through Tidebox.

2. Overage: pay‑as‑you‑go

When credits run out, usage continues at pay‑as‑you‑go rates. No surprise bills — you set a hard cap or auto‑stop per workspace.

3. Hard cap / auto‑stop

Configure a monthly spend limit. When reached, agents pause or fall back to BYOK. You stay in control.

Prefer full control? Use BYOK mode on any plan: plug your own OpenAI / Anthropic / OpenRouter API keys and pay providers directly. Managed credits and BYOK can coexist in the same workspace.
Estimate your inference
20
1
Estimated monthly inference
~€36
Covered by your plan: €25Extra pay‑as‑you‑go: €11

Indicative estimate only, not a quote. Actual costs depend on prompt length, output tokens, and provider pricing.

Add‑ons
Managed inference credit top‑up€10 per €10 block
Dedicated region / serverfrom €199/mo
Break-glass admin access (audited, no public SSH)Enterprise add-on

All prices exclude VAT. Managed inference credits are consumed at provider‑specific rates documented in your billing dashboard.

FAQ

FAQ

🦞

Deploy your first OpenClaw agent in Europe today.

Get a secure EU runtime, connect your tools, and ship an agent with audit logs in minutes.

Founding cohort access
Priority invites for strong fit
Direct founder follow-up
No card required to join
Regulatory posture
GDPR
ALIGNED
NIS2
ALIGNED CONTROLS
SOC 2
READY CONTROLS
EUAI Act
AWARE

These badges describe product posture and control mapping, not a formal certification or legal claim.