Databases that set themselves up
Describe the app; the agent proposes the schema, row-level security and auth, then applies it when you say so.
- Tables
- Row-level security
- Auth
- Indexes
- Types
- Advisors
Global-first data residency
Your data, under the law of the place you choose.
One API key for the leading AI models, in the region you set on every call: resident hosts in Switzerland, the EU and the US, and the cloud that serves each other region. No prompts stored, never trained on.
- Regions
- 10
- PHI regions
- 3
- Providers
- 19
- Prompts stored
- 0
Products
AI gateway
Every leading model, behind one key
Chat, embeddings, speech, OCR and vision, routed to a host in the region you set on each call.
Explore the platform Managed databasesEarly accessPostgres that stays where you put it
Zurich or Frankfurt, set up by an agent that runs in the same jurisdiction.
Ask for early access Health & PHIPatient data, on reviewed hosts only
A hard gate per request: PHI only reaches providers whose DPA or BAA we reviewed.
See health solutionsPowered by the models you already trust
- OpenAI
- Anthropic
- Mistral
- DeepSeek
- Qwen
- Llama
Built to stay put
The model that answers runs in your data’s jurisdiction today. The database that remembers it is next: in early access, in the same jurisdiction.
Your team’s assistant, in your region
Documents, voice and images, every answer routed through the same rules, with the region it ran in printed under it.
10 regions. One key.
Resident providers in Switzerland, the EU and the US carry PHI today; every other region runs general traffic on the clouds that serve it.
- SwitzerlandPHI
PHI: Infomaniak · Microsoft Azure
CH - European UnionPHI
PHI: Microsoft Azure · Scaleway
EU - United StatesPHI
PHI: Microsoft Azure
US - United Kingdom
Google
UK - India
Google
INDIA - Middle East · N. Africa
Google
MENA - Asia-Pacific
Anthropic · Google · Mistral
ASIA - Australia · NZ
Google
ANZ - Latin America
Google
LATAM - Everywhere else
Anthropic · Google · Mistral · OpenAI
WORLD
Take control of where your data lives
Situs is the legal word for where an asset sits, and whose law governs it. CloudSitus makes that a setting: you choose the jurisdiction, and every request honours it or is refused.
Residency on every call
Set the region per request. A request no host there can serve is refused, never moved.
Stateless by design
No prompts, no payloads, no logs of what you send. There is nothing to leak.
Reviewed hosts for PHI
Health data only reaches providers whose DPA or BAA we reviewed, in its own region.
We never train on it
We never train on your data. Health data only reaches hosts whose terms we reviewed; general traffic follows each host’s API terms.
For developers
Keep your client. Change one URL.
Point the SDK you already use at CloudSitus. Two optional headers pin residency and force the PHI tier; the answer comes back with where it ran.
- Chat, embeddings, speech, OCR, vision — one /v1 surface
- x-phi-tier · x-phi-routed · x-phi-attempts on every response
- An MCP server for resellers and agents
# point the client you already use at CloudSitus curl https://cloudsitus.com/api/v1/chat/completions \ -H "Authorization: Bearer $PHI_KEY" \ -H "X-Region: CH" -H "X-PHI: true" \ -d '{"model": "auto", "messages": [{"role": "user", "content": "Summarise this discharge note"}]}'
Response headers
- x-phi-tier
- phi
- x-phi-routed
- infomaniak-nemotron-nano-30b/CH/phi
- x-phi-attempts
- 1
- x-phi-usage
- verified
Pay for calls, not for seats
Prepaid credits
Buy from $10, draw down per call. No subscription, no seats.
Priced per model
One listed price per model, the same at every volume. No PHI surcharge.
Auto-recharge
Top up automatically below a threshold, capped per day.
100 models
Chat, embeddings, speech, OCR and images on one balance.
FAQ
Questions a compliance team asks.
The short, honest answers — the kind that survive a security review.
Do I need a BAA to use CloudSitus?
It depends on your jurisdiction. US PHI for OCR, speech-to-text and text-to-speech runs on Azure under the Microsoft HIPAA BAA; US chat, vision and embeddings PHI are not available. Swiss and EU PHI run under nFADP/GDPR data-processing agreements (Infomaniak/Azure in CH, Scaleway/Azure in the EU) — that is nFADP/GDPR, not HIPAA. The PHI guarantee is the router pinning PHI to an eligible resident provider in your region. Ask us about the right paperwork for your jurisdiction.
Which providers can process PHI?
PHI is region-resident: in Switzerland it routes to Infomaniak or Azure (CH); in the EU to Scaleway or Azure (EU); in the US to Azure (US) for OCR, speech-to-text and text-to-speech only — US chat, vision and embeddings PHI are refused. OpenAI, Anthropic, Google, Mistral and Together AI serve general (non-PHI) traffic only.
Where does my data go?
You pin the region with X-Region (CH, EU, US, …). PHI is filtered to providers eligible in that region; if none are eligible, the request is refused (403 phi_blocked, or 422 no_route where nothing in the region serves that modality) rather than silently crossing a border.
Do you train on my data?
CloudSitus never trains on your data — it is a stateless proxy and stores no prompts or responses. PHI only reaches providers whose DPA or BAA we have reviewed. General traffic is governed by each vendor’s own API terms; where a vendor offers a per-request training opt-out (Deepgram), we send it on every call.
What happens if CloudSitus is breached?
No database holds tenant data (a separate database holds the public model catalogue only), and there is no payload logging — requests authenticate with HMAC-signed self-contained tokens. There is no stored prompt/response content to exfiltrate; usage reaches Stripe as amounts aggregated per model and region.
Will it work with my existing AI code?
Yes — point your existing client at https://cloudsitus.com/api/v1 and it works with no rewrite. Live endpoints cover chat (with image/vision input), embeddings, speech-to-text, text-to-speech, OCR, and /v1/models.
What does it cost?
CloudSitus is prepaid in credits: you buy a credit balance up front (300 credits = $1) and draw it down as you call, at each model's listed per-token price, rescaled to 300 credits/$. You can validate your key and dry-run routing for free with GET /v1/ping before adding credits. No PHI surcharge, no per-seat fees, no subscription. You can pay in your local currency at rounded parity (a $20 pack shows as €20 / £20 / CHF 20); receipts settle in your chosen currency and credits are granted at the USD-parity value.
Are credits the same as the model tokens I see in prices?
No — that's exactly why they have different names. Credits are your prepaid balance, fixed at 300 credits = $1; tokens are what the model counts (the prompt/completion tokens). Model prices on the pricing page are quoted in USD per million model tokens; when you make a call we compute the dollar cost from those model-token counts at the model's listed price and deduct the equivalent credits at 300/$. So 1 credit does not equal 1 model token.
Is there a minimum?
The minimum top-up is $10 (3,000 credits). You can buy a preset pack ($10 = 3,000 credits, $20 = 6,000, $50 = 15,000, or $100 = 30,000) or a custom amount, and there is no subscription or monthly commitment — you only spend the credits you buy. You can pay in your local currency at rounded parity (a $20 pack shows as €20 / £20 / CHF 20), with credits granted at the USD-parity value.
What happens when I run out of credits?
Billable calls return HTTP 402 payment_required until you top up your credit balance in the dashboard → Billing. To avoid interruption, turn on auto-recharge: when your credit balance drops below a threshold you set, we automatically add a fixed amount (e.g. $100 = 30,000 credits) using your saved card.
Ready when you are
Put AI in production — without giving up your data.
One private cloud for the models you already use. Every call runs in the region you choose, we never train on your data or keep your prompts, and every call is yours alone.
Free route checks · Prepaid credits · No data retained