Your AI. Your keys. Your data.
Inferro connects to any OpenAI-compatible endpoint — cloud APIs or your own inference server. Pay once. No subscriptions. No telemetry. Ever.
Built for privacy. Designed for power.
Bring Your Own Keys
Connect to OpenAI, OpenRouter, Mistral, Groq, DeepSeek, Infomaniak, and more — or your own vLLM, Ollama, or LiteLLM. No middleman.
Self-Hosted Inference
First-class support for local and LAN inference. HTTP on LAN and VPN ranges works out of the box. Tailscale, Netbird, WireGuard — all supported.
Zero Telemetry
No analytics, no crash reporters, no ad SDKs. Your chats live in a local SQLite database. Your API keys stay in the iOS Keychain.
Real Agentic Tools
Web search via Brave or SearXNG, URL fetch, project files, model-to-model delegation. Tools run from your device — traffic goes directly to the tools you configure.
Pay Once
One-time purchase. No subscription. No in-app purchases. No "Pro" paywall. The Pro mode toggle is free — it just hides advanced controls for cleaner everyday use.
Made in Switzerland
Built by B&G Digital Services GmbH in Biel/Bienne, Switzerland. We don't take outside investment, we don't run ads, and there's nothing to sell — because we collect no data.
Privacy, actually.
When you send a message, Inferro forwards it directly from your device to your chosen provider. We are not in the path. We see nothing. We store nothing. There is no "our server" — only your device and the provider you chose.
Read the full privacy policy →Ready to take control of your AI?
Inferro is coming soon to the App Store. Built with the privacy and self-hosting community in mind.