Nº 01 Dedicated AI & MCP gateway
Your gateway.Only for you.
No shared endpoint, no API key lost in the crowd: we deploy your AI & MCP gateway dedicated per customer, in our EU cloud or on your premises. With European models out of the box, your tools and your rules.
Jump to Dedicated ↓ AI Gateway ↓ MCP Gateway ↓ Integrations ↓
The gate · Series 01
1:1 Dedicated, not shared
You are not just another API key.
Typical AI gateways are one platform for all customers. With us, every company gets its own gateway, operated where it belongs.
01 AI Gateway
European models, out of the box.
Your gateway comes with European models out of the box and speaks the OpenAI API. Existing tools and SDKs keep working, you only change the base URL.
OpenAI-compatible API
Drop-in for existing SDKs, frameworks and tools like n8n or IDE assistants. Swap base URL and key, done.
European models out of the box
Open-weight models for chat, coding, vision, embeddings and reranking, operated in Europe. Local models on-premise.
Open WebUI included
A chat interface for all employees, connected straight to your gateway. No more shadow AI.
Keys & permissions
Separate keys per team, application or agent, each with its allowed models. Revocable at any time.
Cost & budgets
Usage and cost per model, team and application in real time, with budgets and limits per team.
Switch models without code
A new model is better or cheaper? Switch in the gateway, your applications stay untouched.
import osfrom openai import OpenAI client = OpenAI( base_url="https://<your-gateway>.cloud.apertus.ai/v1", api_key=os.environ["APERTUS_API_KEY"],) reply = client.chat.completions.create( model="deepseek-v4.1-flash", messages=[{ "role": "user", "content": "Summarise the contract in three sentences.", }],)# runs on your dedicated gateway in EuropeWorks the same with cURL, LangChain, LlamaIndex, n8n and every client that speaks the OpenAI API.
EU LLM hosting
LLM hosting in the EU. Ready from day one.
GDPR-compliant LLM hosting without the effort: your gateway starts with a curated selection of current open-weight models operated in Europe. We keep them current and add new models.
- Models from Mistral and Swiss AI from Europe
- Local and dedicated models through the same endpoint
- Closed-source models in the EU region on request
- Mistral AI
- Swiss AI
- DeepSeek
- Qwen (Alibaba)
- Moonshot AI
- Z.ai
- Google DeepMind
- OpenAI
- Meta
- MiniMax
- BAAI
- Your local models
02 MCP Gateway
Your tools. Safely connected.
Agents only become useful once they may access your systems. The MCP gateway is the single, controlled way in: you register MCP servers centrally and decide which agent may use which tool.
Central registry
All MCP servers in one place, reachable for every client through one endpoint.
Permissions per agent
Grants per team, agent and tool instead of full access for every client.
Credentials stay in the gateway
Credentials for your systems are kept centrally, not in prompts, scripts or on laptops.
Every tool call logged
Traceable which agent called which tool, and when.
03 For everyone with a say
One gateway, three pieces of good news.
IT & platform
One endpoint instead of scattered API keys.
Grant and revoke access centrally, swap models without touching applications.
Privacy & compliance
Traceable where data goes.
Operated in Europe, GDPR-compliant, with rules for which data may reach which model and an audit log.
Finance & procurement
Costs you can allocate.
Budgets per team and cost per application, without bulk invoices from several providers.
04 Integrations
One gateway for every tool in your company.
Chat, coding agents or automation: everyone in your company reaches approved models and tools through your gateway, with their own keys, budgets and permissions. Anything that speaks an OpenAI- or Anthropic-compatible API or MCP works.
Coding agents & IDEs
via the Anthropic- or OpenAI-compatible API
- Claude Code
- Codex CLI
- OpenCode
- Cline
- Roo Code
- Kilo Code
- Continue
- Aider
- Zed
Chat & knowledge
for every employee, with SSO and roles
- Open WebUI
- LibreChat
- AnythingLLM
Automation & frameworks
for your workflows and your own agents
- n8n
- LangChain
- LlamaIndex
- OpenAI Agents SDK
- Vercel AI SDK
- Pydantic AI
MCP clients
reach your tools through the MCP gateway
- Claude Code
- Claude Desktop
- VS Code
- Cursor
- Open WebUI
- n8n
# EU model instead of a US cloud, key and budget per personexport ANTHROPIC_BASE_URL=https://<your-gateway>.cloud.apertus.aiexport ANTHROPIC_AUTH_TOKEN=$APERTUS_API_KEYexport ANTHROPIC_MODEL=kimi-k3claude- Personal keys and budgets per person
- Approved models and tools per team
- Costs and usage per tool at a glance
Your tool isn't listed? Almost anything with an OpenAI, Anthropic or MCP interface can be connected.
Contact 20 minutes, an honest assessment
Roll out AI without giving up control.
Book a short call. We look at your situation and tell you honestly whether and how apertus.ai can help.
What happens in the call
- 1You briefly describe where you use AI today.
- 2Using your case, we show what a dedicated gateway, apps and operations look like.
- 3You get an estimate of effort and cost.
- 4Then you decide, not us.