🎙️
LCL Voice
Fast Sub-1.5s Voice Engine
🎨 Image
💬 Text
🎙️ Voice
🔊
Sound Check
📁
History
Fast Cloud
⚙️
🔘
Push-to-Talk
🎙️
Natural VAD
READY
Asteria (Cloud)
STT:
--
LLM:
--
TTS:
--
Total:
--
⚡
Fast Cloud (~1.2s)
💻
Local Mac M1 Max
🎙️
HOLD TO TALK
Press & Speak
Hold button to speak · Release when done
Hold the button to speak or type below.
View Transcript (
0
)
🎙️
🔊
Send
💬
💬
Conversation Transcript
Export TXT
✕
No speech recorded yet. Hold the button to talk or send a text.
📁
Saved Voice Sessions
+ New Session
✕
No past sessions found.
⚙️
Audio & Model Settings
✕
Execution Engine
Auto: Fast Cloud (Deepgram Nova-2 + Grok + Aura) [~1.2s]
Cloud Only (Deepgram + Grok + Aura)
Local Mac Daemon (Whisper + Ollama + Kokoro)
Acoustic Voice Model
🔊
Test Audio
Asteria (Deepgram Cloud Female · Natural)
Luna (Deepgram Cloud Female · Expressive)
Orion (Deepgram Cloud Male · Conversational)
Arcas (Deepgram Cloud Male · Deep)
Nova (OpenAI Cloud Female)
Alloy (OpenAI Cloud Male)
Sarah (Natural Female · US)
Adam (Conversational Male · US)
Bella (Warm Expressive Female · US)
Michael (Deep Male Voice · US)
Emma (British Female · UK)
George (British Male · UK)
Local Brain Model (Ollama)
llama3.2:3b (Instant Sub-500ms · Metal)
gemma4:latest (Ultra-High Quality)
command-r:latest (35B Large Brain)
hermes3:8b (Uncensored Polymath)
qwen3.5:9b (Balanced High-Reasoning)
Conversational Persona
🧠 Polymath
🚀 Co-Founder
⚔️ Sparring
☕ Best Friend
💾
Save Current as Default Settings
✅ Saved as default! These settings will load on every visit.