I'm excited about the Lumo API announced in the autumn/winter 2026 roadmap. As a paying Lumo subscriber and developer working with local AI agent frameworks, I'd like to request one specific design decision:

Please make the Lumo API OpenAI-compatible — i.e., follow the standard chat/completions endpoint format (same request/response structure, streaming support, and authentication headers used by the OpenAI API).

Why this matters: OpenAI-compatibility has become the de facto standard for AI endpoints. Virtually every agent framework, local model router, and developer tool (LangChain, Open WebUI, LM Studio integrations, and countless others) already speaks this format. An OpenAI-compatible Lumo API would be a near-zero-effort drop-in for these tools — users could point their existing agents at Lumo simply by swapping the base URL and adding an API key.
Without this compatibility, each user would need to write a custom adapter, which creates friction and discourages adoption — particularly unfortunate since Lumo's zero-access encryption and no-logging privacy model are exactly what many agent users are looking for in a cloud fallback to their local models.

In short: OpenAI-compatible formatting would maximize the API's reach on day one while preserving everything that makes Lumo unique. Thank you for considering it!