Xinity

Xinity is open-source LLM inference software that runs entirely inside your own infrastructure.

Banks, insurers, hospitals and public sector bodies cannot send prompts, contracts or patient data to a US-hosted API, which leaves most of them either locked out of AI altogether or running unsanctioned tools nobody can audit. Xinity closes that gap. The core is Apache 2.0 licensed and the API is OpenAI-compatible, so existing applications point at a new endpoint and keep working. It deploys on-premise or in a sovereign cloud on your own GPUs, with no outbound calls and no vendor lock-in. Built-in audit trails, role-based access control and model governance give compliance and security teams the evidence they need for GDPR, the EU AI Act and sector-specific supervisory requirements. Built in Vienna for the DACH market, with commercial support, hardware sizing and deployment help from the engineers who wrote the code. Open source so your team can read every line, inspect it, and run it without asking permission.