A self-hosted AI platform, on infrastructure you control.
Run the whole platform — one API for leading models, API keys, teams, usage, and billing — in your own account and region. Operate it yourself, or have us operate it for you.
What runs on your side
Everything your users touch, in an environment you own.
- One API for many models
- Claude, GPT, Gemini, and image models behind a single API, compatible with the SDKs and coding agents your teams already use.
- API keys and accounts
- Organizations, projects, team roles, and API keys, managed in your own console.
- Usage and billing
- Every request metered and priced, with spend tracked per project and per key — for internal chargeback or for billing your own customers.
- Your data, your region
- Accounts, usage records, and billing history stay in your account, in the region you choose.
- Your model catalog
- Choose which models your users can call and what each one costs them.
- Your brand
- Self-hosted works with white labeling, so the platform carries your name, domain, and API key prefix.
Hosted or self-hosted
The same platform either way. The difference is where it runs and who looks after it.
| Feature | Hosted by us | Self-hosted |
|---|---|---|
| Where it runs | Our infrastructure | Your own account and region |
| Who operates it | We do | You do, or we do it for you |
| Account and usage data | Stored by us | Stays in your environment |
| Getting started | Fastest | Depends on your environment |
| Best for | Launching quickly | Data residency, compliance, and control |
How it works
- 01
Tell us about your environment
Where it needs to run, which regions, who will operate it, and what your compliance team needs to see.
- 02
We plan the setup with you
A proposal covering the deployment, the models you want, branding, and how updates reach you.
- 03
Launch on your infrastructure
Your teams or customers get keys and start building, on a platform that lives in your account.
Frequently asked questions
- What is a self-hosted AI platform?
- The software around AI models — the API, keys, accounts, usage tracking, and billing — running in an environment you control instead of a vendor's. Your users get one API for many models; you keep the platform and its records.
- Is this the same as running an LLM locally?
- No. Local tools run one open model on one machine. This is the platform that gives many people access to many leading models, with keys, teams, spend tracking, and billing. If you also run models of your own, tell us when you get in touch.
- Is this a self-hosted alternative to OpenRouter?
- If what you want is one API for many models running in your own account, with your own keys, usage records, and billing, yes — that is what this is.
- Can it run on-premise or in a private cloud?
- Tell us about your environment — a cloud account, a specific region, or your own data center — and we will tell you what we can support.
- Who keeps it up to date?
- We do. New models and improvements are delivered to you, and how they are rolled out is agreed in the setup.
- How is it priced?
- It depends on where it runs, who operates it, and your volume. Get in touch and we will put together a proposal.
Other ways to work with us
- White labelYour own AI platform under your name, logo, and domain.
- Internal AI platformOne AI platform for every team in your company.
- Cloud and telecom providersAdd an AI API to the infrastructure your customers already buy from you.
- AI reseller programResell leading AI models to your customers at your own prices.
Run it on your own infrastructure
Tell us where it needs to run and who will use it, and we will plan a setup that fits.
support@inferray.com