Free & Open Source

Your LLMs,
your keys,
your device.

Arlo Lite is a lightweight iOS client for chatting with LLMs. Bring your own API key and talk directly to the provider — no backend, no subscription, no telemetry.

Arlo Lite app screenshot showing a chat conversation

Built for power users

No fluff. Just a fast, capable interface for working with LLMs every day.

Bring Your Own Key

Your API keys stay on your device in the iOS Keychain. No middleman, no backend server, no account required.

Real-time Streaming

Responses stream in token-by-token via SSE. Stop generation anytime. Provider-specific parsers for maximum fidelity.

Session Management

All chats persisted locally. Switch models mid-session. Rename, delete, or pick up where you left off.

Rich Markdown

Code blocks with syntax highlighting, tables, lists, and one-tap copy. Designed for technical conversations.

Cost Tracking 計劃中

Per-turn and per-session cost computed from token usage and model pricing. Know exactly what you're spending.

Thinking Controls

Adjust reasoning effort per session. Unified abstraction across providers — from minimal to extended thinking.

Multi-provider, one interface

Connect to the models you use. Switch between them seamlessly.

OpenAI

Chat Completions & Responses API. Connect to the latest OpenAI models.

Anthropic

Messages API. Claude Opus, Sonnet, and Haiku models.

Local Inference

Ollama, llama.cpp, vLLM, and any OpenAI-compatible local endpoint.

Privacy by design

No backend server — calls go direct to provider
No telemetry or analytics
API keys stored in iOS Keychain only
No account or sign-up required
Offline read access to past conversations
Optional iCloud backup (text only, no keys)

Open source, always

Arlo Lite is free to use, modify, and distribute. The entire codebase is on GitHub. Built with React Native and Expo for a native iOS experience.

React Native Expo SDK 52 TypeScript Zustand expo-sqlite expo-secure-store
Star on GitHub