Pricing
Pay the provider's rate, plus a flat rate.
Buy credits up front — 300 credits = $1, $20 (6,000 credits) to start — and draw them down as you call. One prepaid credit balance for every model you run, the same provider rate plus a flat +10% margin, and PHI routing carries no surcharge. No subscription, no seats — you spend only the credits you buy.
$ POST /v1/chat/completions X-PHI: true X-Region: CH < HTTP/2 200 < x-phi-tier: phi < x-phi-routed: infomaniak/CH < x-phi-attempts: 1
How pricing works
Two numbers, and nothing hidden behind them.
Buy credits up front (300 credits = $1), then it's a pass-through rate and a flat +10% margin drawn down from your credit balance per call. Here is every part of it, with nothing you have to read the fine print to find.
The provider rate, at the source
You pay each upstream model at its own per-token rate — the regional provider price, not a repackaged flat fee. A model in Switzerland or the EU is often 5–10× cheaper than a US frontier model, and you see exactly which is which.
A flat gateway fee
On top of the provider rate, the gateway adds a flat +10% margin — the only thing phi-cloud charges. No tiers, no enterprise upcharge, the same flat rate at every volume. The prices in the table below already include it.
No surcharge to route PHI
Pinning a request to a PHI-eligible provider costs nothing extra. The PHI tier is priced exactly like any other route — privacy is not a premium.
Buy credits, draw them down
Top up a credit balance through Stripe at 300 credits = $1 — $20 (6,000 credits) minimum, in preset packs ($20 / $50 / $100) or a custom amount. Each call draws down credits at the provider rate plus margin, rescaled to 300 credits/$. 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.
No seats. No monthly commitment.
There are no per-user licenses and no recurring plan. You only ever spend the credits you buy — turn on auto-recharge to top up your credit balance automatically when it runs low, or top up by hand.
Check your setup for free
GET /v1/ping validates your key and dry-runs routing without touching an upstream or spending a credit — so you can confirm auth, region residency and billing status before your first real call.
1M in + 1M out on the PHI tier
Send one million input tokens and receive one million output tokens through Mistral 24B (Infomaniak, CH). At $0.33 in and $0.99 out per million, the whole run lands at the price below — gateway margin already included, PHI routing at no extra cost.
- 1M input tokens
- $0.33
- 1M output tokens
- $0.99
- Total
- $1.32
Live model prices
Every model, every rate, on one page.
Listed per 1M tokens, margin included. Each row shows its region, whether it's PHI-eligible, and a quality tier (Low / Medium / High). Regional rows reflect the local provider price — often 5–10× cheaper than US frontier models.
| Model | Region | PHI | Tier | In / 1M | Out / 1M |
|---|---|---|---|---|---|
Mistral 24B (Infomaniak, CH) infomaniak-mistral-24b | CH | eligible | Low | $0.33 | $0.99 |
Mistral 3 (Infomaniak, CH) infomaniak-mistral-3 | CH | eligible | Low | $0.22 | $0.66 |
Qwen3 (Infomaniak, CH) infomaniak-qwen3 | CH | eligible | Medium | $0.33 | $0.99 |
All-MiniLM-L12-v2 (Infomaniak, CH) infomaniak-minilm-l12-v2 | CH | eligible | Low | $0.02 | — |
BGE-Multilingual-Gemma2 · 3584-d (Infomaniak, CH) infomaniak-bge-gemma2 | CH | eligible | High | $0.06 | — |
Qwen3-Embedding-8B · 4096-d (Infomaniak, CH) infomaniak-qwen3-embedding-8b | CH | eligible | High | $0.09 | — |
Gemma-4 31B IT · vision (Infomaniak, CH) infomaniak-gemma-4-31b | CH | eligible | Medium | $0.24 | $0.48 |
Qwen3.5 397B A17B (Infomaniak, CH) infomaniak-qwen3-5-397b | CH | eligible | High | $0.97 | $4.36 |
Nemotron 3 Nano 30B A3B (Infomaniak, CH) infomaniak-nemotron-nano-30b | CH | eligible | Low | $0.06 | $0.24 |
Panacea 7B · clinical-trials chat (Swiss worker, CH) worker-panacea-7b | CH | general | Low | $0.22 | $0.44 |
Claude Opus 4.7 claude-opus-4-7 | EU | general | High | $16.50 | $82.50 |
Claude Sonnet 4.6 claude-sonnet-4-6 | EU | general | High | $3.30 | $16.50 |
Claude Haiku 4.5 claude-haiku-4-5 | EU | general | Low | $1.10 | $5.50 |
GPT-5 gpt-5 | EU | general | High | $5.50 | $16.50 |
GPT-5 mini gpt-5-mini | EU | general | Low | $0.28 | $2.20 |
Mistral Large 2 mistral-large-2 | EU | general | Medium | $2.20 | $6.60 |
Mistral Small mistral-small | EU | general | Low | $0.22 | $0.66 |
Gemini 2.5 Pro gemini-2-5-pro | EU | general | High | $1.38 | $11.00 |
Gemini 2.5 Flash gemini-2-5-flash | EU | general | Low | $0.33 | $2.75 |
Gemma-4 26B IT · vision (Scaleway, EU) scaleway-gemma-4-26b | EU | eligible | Medium | $0.30 | $0.59 |
Qwen3.6 35B A3B · vision (Scaleway, EU) scaleway-qwen3-6-35b | EU | eligible | Medium | $0.48 | $1.78 |
Devstral-2 123B Instruct (Scaleway, EU) scaleway-devstral-2-123b | EU | eligible | Medium | $0.48 | $2.38 |
Qwen3-Coder 30B A3B Instruct (Scaleway, EU) scaleway-qwen3-coder-30b | EU | eligible | Low | $0.24 | $0.95 |
Qwen3.5 397B A17B · math/reasoning (Scaleway, EU) scaleway-qwen3-5-397b | EU | eligible | High | $0.71 | $4.28 |
Qwen3-Embedding-8B · 4096-d (Scaleway, EU) scaleway-qwen3-embedding-8b | EU | eligible | High | $0.12 | — |
Claude Opus 4.7 claude-opus-4-7 | US | general | High | $16.50 | $82.50 |
Claude Sonnet 4.6 claude-sonnet-4-6 | US | general | High | $3.30 | $16.50 |
Claude Haiku 4.5 claude-haiku-4-5 | US | general | Low | $1.10 | $5.50 |
GPT-5 gpt-5 | US | general | High | $5.50 | $16.50 |
GPT-5 mini gpt-5-mini | US | general | Low | $0.28 | $2.20 |
Gemini 2.5 Pro gemini-2-5-pro | US | general | High | $1.38 | $11.00 |
Gemini 2.5 Flash gemini-2-5-flash | US | general | Low | $0.33 | $2.75 |
Qwen3.5 397B A17B · math/reasoning (Together, US) together-qwen3-5-397b | US | general | Medium | $0.77 | $3.85 |
GLM-5.2 · chat + code (Together, US) together-glm-5-2 | US | general | High | $0.66 | $2.41 |
GLM-5.2 · chat + code (Z.ai native) zai-glm-5-2 | US | general | High | $0.66 | $2.42 |
Kimi K3 · chat + code + vision (Moonshot native) kimi-k3 | US | general | High | $3.85 | $16.50 |
Prices are USD per 1M model tokens (the prompt/completion tokens a model counts), gateway margin included — not to be confused with credits, the prepaid balance your wallet holds at 300 credits = $1. Credits are what you buy; tokens are what the model counts. These are the live CH / EU / US token models; audio, OCR and image/video price per minute / page / image / second. The same rows — with quality_tier and phi_eligible per model — are served live at GET /v1/pricing, and the full catalog (all regions + capabilities) at GET /v1/models.
FAQ
Questions a compliance team asks.
The short, honest answers — the kind that survive a security review.
It depends on your jurisdiction. US PHI runs on Azure under a signed Microsoft HIPAA BAA. 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.
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). OpenAI, Anthropic, Google, Mistral and Together AI serve general (non-PHI) traffic only.
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 (a 403 phi_blocked) rather than silently crossing a border.
No. phi-cloud is a stateless proxy — it stores nothing about your traffic and never trains on it. Upstream providers are bound by their own terms; PHI-eligible providers are selected for non-training/DPA terms.
There is no application database and no payload logging — requests authenticate with HMAC-signed self-contained tokens. There is no stored prompt/response content to exfiltrate; the only billing surface is per-call token counts in Stripe.
Yes — point your existing client at https://phi-cloud.com/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.
phi-cloud is prepaid in credits: you buy a credit balance up front (300 credits = $1) and draw it down as you call, at the upstream provider's per-token rate plus a flat +10% gateway margin, rescaled to 300 credits/$. 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.
No — that's exactly why they have different names. Credits are your prepaid balance, fixed at 300 credits = $1; tokens are what the model counts (the prompt/completion tokens). Model prices on the pricing page are quoted in USD per million model tokens; when you make a call we compute the dollar cost from those model-token counts (provider rate + 10% margin) and deduct the equivalent credits at 300/$. So 1 credit does not equal 1 model token.
The minimum top-up is $20 (6,000 credits). You can buy a preset pack ($20 = 6,000 credits, $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.
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. Your data stays in your region, never trained on, every call yours alone.
Free to test · Prepaid credits, no subscription · No data retained