Documentation

Lorivo hosts your LoRA adapters on a GPU node and serves them through an OpenAI-compatible API. Use the CLI for the fastest path, or raw REST requests when you want full control.

Quickstart

Install the CLI, deploy your first adapter, and call it in about five minutes.

CLI reference

Every lorivo command (login, deploy, list, rollback, keys) with flags and environment variables.

REST API

Raw HTTP for the OpenAI-compatible inference endpoints and the adapter lifecycle.

Two ways to use lorivo

  • CLI: lorivo deploy validates, uploads, and deploys an adapter in one command, then prints your inference endpoint.
  • REST API: call /v1/chat/completions, /v1/responses, and the management endpoints directly with an API key, from any HTTP client or the OpenAI SDKs.

New here? Start with the Quickstart.