VoiceLM

Talk to your local LLM — with your voice.

VoiceLM connects to LLM servers running on your local network (LM Studio, Ollama) and lets you have natural voice conversations with AI.

VOICE CONVERSATION
- Speak naturally — Apple Speech Recognition converts your voice to text
- AI responses are read aloud with text-to-speech
- Hands-free mode: automatic silence detection sends your message, listens again after the AI responds

LOCAL & PRIVATE
- All processing stays on your network — no data leaves your device
- Voice recognition and text-to-speech use built-in iOS engines — no external services needed
- No account required, no tracking, no ads
- Connect to any OpenAI-compatible local LLM server

BUILT-IN AI MODE
- On-device AI powered by Apple Intelligence — no server needed
- Requires Apple Intelligence supported device (iPhone 15 Pro / Pro Max or later,　iOS 18.1+)
- Siri language must match your iPhone's language setting to enable Apple Intelligence

CUSTOMIZABLE
- 3 server presets — switch between models instantly
- Leave model name empty to use whichever model is currently loaded on your server
- Adjustable LLM parameters: Temperature, Max Tokens, Context History
- System prompt support — customize AI behavior
- 10 languages supported for voice recognition and speech
- 7 theme colors

Simple, private, and powerful. Your AI, your voice, your network.

Built in collaboration with AI — an app made through human-AI dialogue, for human-AI dialogue.

REQUIREMENTS
- Local LLM Mode: Requires an LLM server (Ollama, LM Studio, etc.) running on your local network
- Built-in AI Mode: Requires Apple Intelligence (iPhone 15 Pro / Pro Max or later,　iOS 18.1+)
- Siri language must match your iPhone's language setting to enable Apple Intelligence
- Bluetooth audio (especially car HFP) may cause voice instability