FAQ
What is Nexis?
Section titled “What is Nexis?”An open-source, lightweight, cross-platform AI-native terminal and developer environment. Terminal, editor, language tooling, source control, and an AI panel in one app. See What is Nexis?.
What is it built with?
Section titled “What is it built with?”Tauri 2, Rust, and React 19. The native PTY and heavy lifting are in Rust; the UI renders in the system webview. Nexis is designed to stay responsive and native, but no longer sacrifices useful capabilities to a fixed size target.
Is it free and open source?
Section titled “Is it free and open source?”Yes — Nexis is released under the Apache-2.0 license. It builds on the open-source terax-ai project.
Does Nexis send my data anywhere?
Section titled “Does Nexis send my data anywhere?”Nexis does not include usage telemetry. Code stays local unless you choose to send context to a cloud AI provider or use another networked feature. API keys are stored through the OS keychain, not in Nexis settings files. The keychain may persist encrypted data on disk. The AI context inspector shows what context is sent to a model.
Can I use it without any AI provider / API key?
Section titled “Can I use it without any AI provider / API key?”Yes. Point Nexis at a local LM Studio, MLX, or Ollama endpoint and it runs fully offline with no key. See AI providers.
Which AI providers are supported?
Section titled “Which AI providers are supported?”12+, including OpenAI, Anthropic, Google, Groq, xAI, Cerebras, DeepSeek, Mistral, OpenRouter, Hugging Face, any OpenAI-compatible endpoint, and the local runners above. Full list on AI providers.
Will the AI edit or run things without asking?
Section titled “Will the AI edit or run things without asking?”By default, tools prompt before they run. Policies can also deny a tool, allow it, or auto-approve only shell commands that pass a strict read-only check. AI-proposed edits appear as per-hunk diffs, and agent checkpoints provide a one-click Git-backed revert path.
What platforms are supported?
Section titled “What platforms are supported?”macOS, Linux (.deb / .rpm / AppImage / AUR), and Windows (NSIS / MSI), with
first-class WSL. See Installation.
What is the ML Lab?
Section titled “What is the ML Lab?”A panel for training small models on your own data, powered by a local engine —
either the Python nexis-ml or the single-binary Rust
nexis-ml-rs. See ML Suite.
How do I report a bug or request a feature?
Section titled “How do I report a bug or request a feature?”Open an issue on the Nexis GitHub repository.