Docs
Nexora is a local-first AI desktop workspace. These docs explain how to set it up, connect providers, run agents, and where to find the specification documents in the repository.
What these docs cover
- Setting up Nexora: prerequisites, dev build, and production bundle.
- Connecting AI providers: OpenAI, Anthropic, Google Gemini, OpenRouter, NVIDIA NIM, OpenCode Zen, or your own OpenAI-compatible endpoint.
- Agent runs: autonomy levels, approval gates, and run controls.
- Where your data lives: single SQLite file on your disk, OS keyring for credentials.
- Links to the specification documents and the changelog.
Audience
These docs are for people running Nexora on their own machine. There is no cloud account, no sync service, and no multi-user portal. If you are looking for the backend API reference, Nexora is a Tauri desktop app with local IPC commands - it does not expose a public HTTP API.
Specification documents
Approved specification documents live in thedocs/ directory of the repository:
Provider setup
Nexora connects to the providers you configure yourself. Credentials are stored in the OS keyring and are never written to SQLite, logs, or error text. Supported providers:
- OpenAI
- Anthropic
- Google Gemini
- OpenRouter
- NVIDIA NIM
- OpenCode Zen
- Custom OpenAI-compatible endpoint