visibus Chat - one AI interface for your entire team.
An isolated instance under your domain, all the major AI models in the same chat, a privacy filter up front, budget per team. We set it up, you stay in control.
Alle führenden KI-Modelle - inklusive lokaler Modelle. Eine Plattform.
Everything your company needs for AI
All models. No dependency.
Your team picks the best model for each task - we provide the selection.
- Claude, GPT, Gemini, Mistral, Llama, DeepSeek
- New models enabled within a day
- No vendor lock-in - switch models anytime
Privacy that deserves the name.
An isolated instance in Switzerland or Germany - GDPR- and FADP-compliant.
- Strict separation from other customers
- Privacy filter anonymizes sensitive data before the AI model
- DPA, TOMs, and sub-processor list prepared
- No provider trains on your data
Input
Contract with Hans Muster, IBAN CH93 0076...
To AI
Contract with <PERSON>, IBAN <IBAN>
Full control over cost and usage.
You see in real time who uses what and what it costs.
- Team budget as a spending limit
- Usage overview per user and model
- Alert when a user exceeds their limit
- Enable or block models, manage access rights
- Usage-based billing instead of a flat rate
Managed service. Everything included.
We run the platform for you - from setup to support.
- Setup, updates, and monitoring
- Direct support, no ticket system
- Onboarding in 30 minutes, productive from day 1
- Also available as a self-hosted option
And on top of that
Here's what it looks like in practice.
One browser tab, all models. Switch in seconds - no new login, no new tab.
Which tool do I use for this question?
Which login was that again? Who has the premium subscription?
Sensitive data? Stays sensitive.
The privacy filter detects personal data automatically and anonymizes it - before it reaches the AI model. The AI answers based on anonymized data, while you see the original.
Summarize the contract with Hans Muster for me.
Address: Bahnhofstrasse 42, 8001 Zürich
IBAN: CH93 0076 2011 6238 5295 7
AHV no.: 756.1234.5678.90
Summarize the contract with <PERSON> for me.
Address: <ADDRESS>
IBAN: <IBAN_CODE>
AHV no.: <AHV_NUMBER>
Automatically. In real time. Without any action from your employees.
The employee sees the real name and the real IBAN. The AI works only with the anonymized placeholders - and still delivers a correct, usable draft.
Your company knowledge. Always available.
Documents are more than attachments. Depending on the use case, there are four ways to bring knowledge into visibus Chat - from a one-time upload to permanent company memory.
File upload in the chat
For one-off analyses
Upload a file directly into the chat - contract, quote, report, spreadsheet. The AI reads and analyzes the content on the spot. No permanent storage needed.
Example
Upload a contract → "Which clauses are unusual or risky?"
Folders & projects
For ongoing topics
Create folders for recurring topics - a customer project, a tender, the annual budget. All related documents stay bundled together. The context is right there in every conversation.
Example
Folder "Müller Construction Project" with plans, minutes, and quotes → pick up right where you left off
Knowledge base
For permanent company knowledge
Store documents that should always be available: product catalog, staff regulations, internal policies, price lists. Reference them in the chat with "#" - the AI answers based on your documents, not the internet.
Example
"Create a quote for Müller AG, use #ProductCatalog-2026"
Notes
For templates & writing guidelines
Short instructions that should always apply: writing style, email template, tone, standard phrasing. The AI reads the note automatically - you never have to explain your guidelines twice.
Example
Note "Company style: formal tone, max. 3 sentences, no anglicisms" → consistent copy without a briefing
What happens to your data?
In your instance
Documents, search, and context building run on your server in Germany or Switzerland. Encrypted transmission, no external indexing services.
At the AI provider
When you ask a question, the relevant excerpts go to the chosen model. In their API terms, the providers used commit to no training on your data and only short-term processing - unlike free consumer versions.
When that's not enough
Available as an add-on: the PII filter anonymizes names, addresses, IBANs, and social security numbers before the request leaves your instance - the provider sees only placeholders.
What your team does with it
Real scenarios from everyday work - by department. Where your team spends hours on repetitive text work today, it's often minutes with visibus Chat. We'll work out how big the benefit is per task for your business in the free AI check.
Questions about internal policies, staff regulations, or labor law - the AI answers based on your own documents, not the internet.
What does #StaffRegulations say about working from home? Does it also apply to part-time staff?
Per §12 of your staff regulations: - Max. 2 days working from home per week - Applies to all employment levels from 50% up - Requires supervisor approval - Reachable during core hours (09:00–16:00) Would you like to see the full section?
You describe the task. visibus Auto picks the model.
No guessing which model to use: you ask your question, visibus Auto picks the right model. And you always see which one answered.
"Analyze the risks from the quarterly report and summarize the three most important points."
Automatically the right fit
visibus Auto recognizes what's needed - coding, analysis, translation, long documents - and takes the model that fits best. No one has to choose.
Choose for yourself anytime
Anyone who prefers to decide can pick a specific model directly - such as Claude, ChatGPT, or a European model. Full control stays with your team.
Transparent
Under every answer, it shows which model responded - fully traceable.
For sensitive data, the privacy filter kicks in as well: personal data is anonymized before it reaches any model. If you like, you can deliberately choose a European model.
Which model does what?
Not every AI model is equally good at every task. With visibus Chat, your team always picks the right one.
Which task - which model?
As of April 2026| Task | Recommended model | Why |
|---|---|---|
| Code & scripts | Claude Sonnet 4.5+ | Best performance for snippets, SQL, scripts, and coding help |
| Multi-step analysis, math, logic | Claude / OpenAI o-models (thinking mode) | Models that reason internally before answering |
| Very long documents | Gemini 2.5 / Claude (1M context) | Longest context window - entire contracts or reports at once |
| Fast, low-cost routine tasks | Gemini Flash, Claude Haiku, GPT-4o-mini | Low latency, low cost per request |
| Image understanding | Claude, GPT-4o, Gemini (on par) | All three deliver strong multimodal results |
| EU hosting, French | Mistral Large | EU provider, strong in Romance languages |
| Confidential data | Local model (Mistral, Llama via Ollama) | Never leaves your infrastructure |
visibus Chat is the universal AI platform - the digital Swiss Army knife for your company. For highly specialized workflows like integrated software development, there are dedicated tools. But when it comes to giving your entire team controlled access to AI, visibus Chat is the ideal choice.
| Capability | Claude Sonnet | GPT-4.1 | Gemini 2.5 | Mistral Large | Llama 4 | DeepSeek R1 |
|---|---|---|---|---|---|---|
| Text & Chat | ||||||
| Code | ||||||
| Image understanding | ||||||
| Documents (PDF) | ||||||
| Long context | ||||||
| Reasoning | ||||||
| Speed | ||||||
| Value for money |
Tap a model to see details.
All available models
Anthropic
Claude Opus 4.6, Claude Sonnet 4.6, Claude Haiku 4.5
OpenAI
GPT-4.1, GPT-4.1 mini, GPT-5, o3, o4-mini
Gemini 2.5 Pro, Gemini 2.5 Flash
Mistral
Mistral Large, Mistral Small
Meta
Llama 4 Maverick, Llama 4 Scout
DeepSeek
DeepSeek R1, DeepSeek V3
xAI
Grok
Cohere
Command R+
Perplexity
Sonar Pro, Sonar
More models on request. Local models available via Ollama.
New models are added continuously. You decide which ones your team may use.
Platform features - model-independent
These are the features visibus provides around the models. They work with any provider - you switch the model, the features stay.
Web search
Up-to-date information from the web - works with any model.
Knowledge bases
Bring in your own documents; the model answers from your company knowledge.
Privacy filter
Sensitive data in chat requests and knowledge content is filtered before it reaches the model.
Budget control
Limits per team and per user, with a transparent real-time overview.
Audit log
Who used what and when - fully traceable.
Roles & permissions
Who can see which model - configurable per team.
How you manage models for your team
In the admin tool, you enable models per team or plan. Capabilities, context lengths, and price tiers are clearly documented.
visibus Chat vs. the alternatives
| visibus Chat | Copilot (M365) | ChatGPT Business | Claude Team | |
|---|---|---|---|---|
| AI models | All providers | GPT only | GPT only | Claude only |
| Data location | CH / DE (isolated) | Microsoft Cloud | US Cloud | AWS (US/EU) |
| Privacy filter | Two-stage | |||
| Vendor lock-in | None | Microsoft | OpenAI | Anthropic |
| Pricing model | Team budget + per-user limit | Fixed price per user | Fixed price per user | Fixed price per user |
| Setup | Managed, a few days | Self-service | Self-service | Self-service |
| Support | Personal (CH) | Ticket system | Ticket system | Ticket system |
Copilot is strong in M365 integration (Teams, Excel, PowerPoint). visibus Chat is the better choice for general AI use with a privacy focus.
What fits your company?
Privacy doesn't carry the same weight in every company. visibus Chat covers all the industries shown here - the difference is in the configuration, not the solution.
Standard SME
Consulting, retail, industry, agencies
What you need
DPA, TOMs, EU/CH hosting - the GDPR baseline.
What we deliver
A complete package as standard. DPA, TOMs, and sub-processor list are tailored to your company during onboarding.
What this looks like day to day
A retail company has its team draft quotes, customer emails, and product copy in visibus Chat - in the browser, with all leading models, instead of private AI subscriptions. The DPA and sub-processor list are in place from day one.
Health data
Practices, clinics, rehab, HR handling diagnoses
What you need
Pseudonymization before the AI model, input for your data protection impact assessment (DPIA).
What we deliver
Extended privacy filter with detection of free-form names and addresses. TOM annex as DPIA input. Filter sensitivity adjustable per instance.
What this looks like day to day
A medical practice prepares reports and patient correspondence with visibus Chat. The extended privacy filter detects names, addresses, and diagnoses and replaces them before the request reaches the model. The TOM annex serves as input for the data protection impact assessment.
Professional secrecy
Law firms, notaries, tax advisors, medical practices
What you need
Client and patient data must not reach cloud providers (§203 StGB / Art. 321 StGB).
What we deliver
Extended filter, self-host option, and a local model on request. Plainly put: no provider can waive professional secrecy - the filter reduces the risk that confidential content reaches the cloud provider. For the highest sensitivity, a local model is the clean answer.
What this looks like day to day
A tax advisory firm uses visibus Chat to prepare client correspondence and receipts faster. The privacy filter replaces names, addresses, and amounts before the request reaches the model. For especially sensitive content, the firm works with a local model - the data never leaves its own instance.
Banking, wealth management
Swiss banks, wealth managers
What you need
Banking secrecy (Art. 47 BankG) - customer data outside your own house is sensitive.
What we deliver
Self-hosting on your infrastructure, plus a local AI model on request. No external provider, no third-country transfer.
What this looks like day to day
A wealth manager runs visibus Chat on its own infrastructure with a local model. Research, analysis, and drafts are created in house, without customer data going to an external provider.
"No US cloud" policy
Corporations, government agencies, public sector
What you need
Processing exclusively in the EU/CH or fully local.
What we deliver
A choice between the EU provider Mistral (France) and a local model on request. EU/CH hosting as standard.
What this looks like day to day
A government agency uses visibus Chat with the EU provider Mistral and a local model. The instance runs on EU/CH hosting, with no US services involved.
Swiss companies (revised FADP)
Swiss SMEs, associations, NGOs
What you need
DPA under Art. 9 of the revised FADP; CH data residency often desired.
What we deliver
CH hosting available, DPA under Art. 9 of the revised FADP. From a Swiss perspective, EU providers are uncritical (the EU is on the list of countries with adequate protection).
What this looks like day to day
A Swiss SME uses visibus Chat with CH hosting for its entire team. The DPA under Art. 9 of the revised FADP is in place, and the instance is strictly separated from other customers.
All the industries shown can be covered with visibus Chat - the difference is the configuration, not whether it's possible.
Don't see your situation here? In the AI check we discuss your specific requirements - GDPR, the revised FADP, industry regulations, or internal corporate policy.
These scenarios show typical use cases - not named customers. We're happy to discuss concrete references from your industry in the AI check, once we're allowed to share them with the respective company's consent.
Custom terms for your company
From CHF 290/month
Team bundle - no per-user license fees.
Included in every plan
- Hosting in CH/DE
- All leading AI models
- Privacy filter (basic included, extended optional)
- Personal support from Switzerland
- Onboarding and training
- Updates and maintenance
Real-world examples
Small team
- • up to 10 users
- • DE hosting
from CHF 290/month
Mid-size team
- • up to 25 users
- • extended privacy
from CHF 450/month
Larger team
- • up to 50 users
- • extended privacy
- • CH hosting
from CHF 730/month
Tailored
- • more than 50 users
- • custom quote
on request
The prices shown are the fixed monthly fee. AI usage (LLM requests) is billed separately - usage-based, depending on the model and usage. With your own API keys, you're billed directly by the provider with no markup; otherwise a 20% pass-through markup applies. Annual billing gets a 10% discount on fixed items. Exact quote after your setup - typically within 1–2 days.
Prefer to self-host?
visibus Chat on your own infrastructure.
Quote on request
Your own AI platform in 5 steps
Not an IT project that drags on for months. We handle the setup, you set the pace.
AI check
Which models? What level of privacy? How many seats? In a free conversation, we clarify the scope - you get a concrete assessment of where AI brings the most value for you, and a clear recommendation.
Infrastructure setup
We set up the server and AI stack - in a few days, with nothing required from your IT.
Trial run with a pilot group
Before the whole team is switched on, a small group tests your instance in everyday work. That way you see the real benefit before you roll out broadly - keeping the risk small. The feedback goes straight into fine-tuning.
Rollout & onboarding
All users are switched on, admins are trained, and employees get a compact introduction.
Ongoing support
Updates, monitoring, support - and we enable new models as soon as they're available.
AI check
Which models? What level of privacy? How many seats? In a free conversation, we clarify the scope - you get a concrete assessment of where AI brings the most value for you, and a clear recommendation.
Infrastructure setup
We set up the server and AI stack - in a few days, with nothing required from your IT.
Trial run with a pilot group
Before the whole team is switched on, a small group tests your instance in everyday work. That way you see the real benefit before you roll out broadly - keeping the risk small. The feedback goes straight into fine-tuning.
Rollout & onboarding
All users are switched on, admins are trained, and employees get a compact introduction.
Ongoing support
Updates, monitoring, support - and we enable new models as soon as they're available.
Frequently asked questions
Let's talk about your AI strategy.
In the free AI check, we show you what visibus Chat would look like in your company - or you test the platform directly for 7 days in your everyday work.
Free and non-binding. An open conversation, no sales pressure.