Infercom Solution Partner · data centres in Germany

AI for your business that stays in Europe.

Questions, documents, research and verified law: Suveri answers everything with open models on Infercom in German data centres. No storage by third parties, every answer with model, cost and sources.

EU data centres, zero data retentionEvidenced answers with VerifactCost per answer visibleAlso available in Claude and ChatGPT

Everything teams need every day

One workspace for questions, documents and research, on open models, with transparent cost.

Understands documents

PDF (including scans), Word, Excel, PowerPoint, images, audio and video. Large files become the conversation's knowledge base, with sources per answer.

Live research

Our server reads links itself; research searches via our own meta search and reads the best hits. Sources and dates appear under every answer.

Verified law

Verifact provides German law with official references and a trust score as the evidence base, switchable per conversation.

Pick a model or let the router choose

MiniMax, gpt-oss, Gemma, Whisper and more. The router chooses per task, or you fix the model. The price per model sits right next to it.

Quality levels and Challenge me

Verified: a second model flags what is unsupported. Ensemble: two models answer, a third merges. Challenge me lets senior roles disagree.

Protection against hidden instructions

Invisible text in files and web pages is detected and removed, links in answers are checked. Measured in our red-team eval: 0 of 130 attacks followed.

EU AI Act, under control

The AI register keeps your use cases, classifies them under the EU AI Act (questionnaire → risk tier), lists the obligations, records approvals and exports an audit trail as PDF.

Sovereignty and security

Why Suveri is the more credible route to EU AI – demonstrable, not just promised.

ISO 27001 certified

Operated by Pioneerdesk GmbH – certified to ISO/IEC 27001 for information security. Audited processes, not self-assessment.

Self-hosted, not a proxy

We run open models ourselves on STACKIT in Germany. Your data is not forwarded to US providers – no CLOUD Act residual risk.

Zero data retention

Prompt, files and answer are never used for training or stored permanently. History encrypted, deletable at any time.

Built-in data-protection filter

Optional guardrails detect and redact personal data and prompt injection before they reach a model – rule-based and local.

Automatic EU fallback

If a model fails, the router switches to another EU model on its own – never to global infrastructure. Requests stay sovereign and keep running.

How it works

Three steps, nothing to install.

Register

Ask or upload

Write in the chat, drop files, switch on Verifact or research. The router picks the right model.

Pay for what you use

Top up credit; every answer shows tokens and cost. Companies get allowances, beyond that the token price applies.

Suveri in everyday work

Four examples — just type the question or drop in a document.

Check a contract

„Check this contract for risks and deadlines.“
→ You get a risk table with real citations (court, file no., date) and the key clauses.

Compare offers

„Compare these two offers (Excel).“
→ A table of total cost, differences in scope and the better offer — with reasoning.

Meeting minutes

„Turn this recording into minutes with to-dos.“
→ Decisions, tasks (who, what, by when) and open questions — as clean minutes to download.

Research dossier

„Research the market and build an investor dossier.“
→ A structured document with summary, sections and sources — every claim with a date.

And beyond that

Suveri is your sovereign AI workspace. These extras bring it to where you already work.

In Claude, ChatGPT and Gemini

As an MCP connector you use the same account and the same models right inside your AI assistant.

In the terminal, with coding routing

One command, browser login instead of tokens: “suveri-code login”, then you code with aider or any OpenAI-compatible tool against EU models. Routing picks the right model by alias, and afterwards the CLI starts your local runtime for testing. Same account, billing and data-protection filters as in chat.

Your knowledge and your Git

Files, whole folders and repositories from GitHub, GitLab or STACKIT Git flow into your knowledge base, encrypted and tenant-separated.

Skill packs

Curated ways of working: law with verified citations, analysis and research, and a tech team from intake to architecture.

Also inside your assistant

Suveri is a connector too: Claude, ChatGPT and Gemini CLI delegate tasks to the same account, with the same billing.

Claude Desktop, Cowork, claude.ai, mobile (Pro, Max, Team, Enterprise; Free with one connector)

Customize → Connectors → Add custom connector.
Enter the https://suveri.de/mcp: https://suveri.de/mcp. No other fields.
Claude opens our sign-in. Register or sign in, done.
In chat: “Use Infercom and summarise this document.” Claude calls the tool ask.

Team/Enterprise: an owner adds the connector in organisation settings, members enable it themselves. Claude Code: claude mcp add --transport http suveri https://suveri.de/mcp

ChatGPT (Plus, Pro, Business, Enterprise, Edu; web app)

Settings → Apps → Advanced settings → enable Developer mode. In Business/Enterprise an admin has to allow it.
Apps → Add connector → https://suveri.de/mcp https://suveri.de/mcp, authentication OAuth.
Complete the sign-in and enable the connector in chat.
Use it in chat or Deep Research. For Deep Research, search and fetch are available.

Developers: via the Responses API as a tool of type mcp with the same URL.

Gemini

The Gemini app has no custom connectors yet. The router works there through the Gemini CLI, which supports MCP servers:

Add to ~/.gemini/settings.json under mcpServers: {"suveri": {"httpUrl": "https://suveri.de/mcp"}}
Start Gemini CLI; /mcp lists the server and sign-in opens in the browser.
Phrase tasks as you would in Claude.

Once Google enables connectors in the Gemini app, the same endpoint works unchanged.

Pricing

Usage-based per token, net. No subscription required. With your own Infercom key, Infercom bills you directly. For example, €25 of credit buys about 46.175.824 output tokens on gemma-4-31B-it.

Model	Region	Capabilities	Context	Input / 1M	Output / 1M
`gemma-4-31B-it`	EU	vision	131.072	0.26 €	0.46 €
`gpt-oss-120b`	EU	reasoning, tools	131.072	0.29 €	0.77 €
`MiniMax-M2.7`	EU	tools	196.608	0.78 €	3.12 €
`Meta-Llama-3.3-70B-Instruct`	Global	text	131.072	0.78 €	1.56 €
`DeepSeek-V3.1`	Global	reasoning, tools	131.072	3.90 €	5.85 €
`DeepSeek-V3.2`	Global	reasoning	32.768	3.90 €	5.85 €
`E5-Mistral-7B-Instruct`	EU	embedding	4k	0.17 €	–
`Whisper-Large-v3`	EU	audio	–	0.13 € / hour	–

All 8 models of the Infercom catalogue (synced daily with api.infercom.ai). EU = data centres in Germany, the router's default. Global = infrastructure in the US/Japan, only on explicit request. Prices net of VAT; Whisper per audio hour, E5 input tokens only. Top-up from €25 gross, invoice by e-mail immediately.

Optional: a Verifact answer in chat costs 0.10 € per question, net, only while the switch is on and Verifact returns an evidenced answer. Free for Infercom staff.

Frequently asked questions

Answered briefly and honestly.

What is Suveri?

A sovereign AI workspace for businesses, from the EU: chat, documents, research, a knowledge base and curated skills, on open models, with transparent cost.

Where is my data processed?

EU-sovereign, in data centres in Germany. Inference runs with zero data retention, meaning your inputs and outputs are not stored and not used for training.

Which AI models does Suveri use?

Open EU models such as MiniMax, gpt-oss and Gemma, plus Whisper for speech and E5 for embeddings. Global models like DeepSeek only if you explicitly choose them, and an operator can block them entirely; otherwise the router picks the right one.

What does Suveri cost?

Your first 20 requests per business account are free – registration requires a business e-mail (no free or disposable addresses). After that you pay from credit based on usage, or via a subscription. Per-model prices are shown transparently in the portal.

Do I need my own Infercom key?

No. By default everything runs through Suveri. If you have your own Infercom key you can add it and pay Infercom directly for the models, plus a small platform fee.

Can I use Suveri in Claude, ChatGPT or the terminal?

Yes. As an MCP connector in Claude, ChatGPT and Gemini, and via an OpenAI-compatible endpoint in terminal tools like aider or opencode.

Can I connect my files and Git repository?

Yes. Single files, whole folders via browser or CLI, and repositories from GitHub, GitLab or STACKIT Git. All encrypted, separated per organisation, included in knowledge via a toggle.

What are skills and skill packs?

Curated ways of working on a topic. They give the answer first, then a full-circle view from all relevant angles, and actively close gaps. Legal skills cite real sources from Verifact.

How secure and GDPR-compliant is Suveri?

Data encrypted, strict tenant separation, protection against prompt injection, and an architecture reviewed against an internal red team. The operator is a German company with a data-processing agreement.

Who operates Suveri?

Pioneerdesk GmbH as an Infercom solution partner. You will find the imprint and legal texts in the page footer.

Does Suveri help with the EU AI Act?

Yes. The AI register records your AI use cases, suggests a risk tier under Regulation (EU) 2024/1689 via a questionnaire, lists the associated obligations, provides an approval workflow and exports an audit trail as PDF. Not a substitute for legal advice.

Start with your first question.

Registration takes a minute – business e-mail only.