⚙️ Cogs Settings

Configure AI services, voice, and API settings
🎤 Voice & Speech
Checking...
PersonaPlex is full-duplex voice AI (ASR+LLM+TTS in one, 70ms latency). Kokoro is high-quality local TTS. Browser Web Speech works on all devices.
Bella
Warm female (default)
Alloy
Clear female
Aoede
Melodic female
Heart
Friendly female
Jessica
Professional female
Kore
Youthful female
Nicole
Confident female
Nova
Bright female
River
Calm female
Sarah
Natural female
Sky
Airy female
Adam
Deep male
Echo
Resonant male
Eric
Warm male
Fenrir
Strong male
Liam
Friendly male
Michael
Professional male
Onyx
Authoritative male
Puck
Playful male
Alice
British female
Emma
British female
Isabella
British female
Lily
British female
Daniel
British male
Fable
British male
George
British male
Lewis
British male
27 high-quality neural voices, runs locally on GPU - zero API cost
🎤 Voice Cloning
Clone a voice from an audio sample using ElevenLabs. Record or upload a voice sample (at least 30 seconds recommended).
— or record directly —
🪪 Identity
Give your AI a custom name (used in chat and prompts)
📞 Phone Voice
This voice is used when Cogs speaks to you on the phone via Twilio. Changes are saved to your profile.
Optional instructions that shape how Cogs responds to you (applies to chat, phone, and SMS)
🎨 Face Theme
Classic Blue
Warm Amber
Neon Green
Soft Pastel
Dark Crimson
Changes apply instantly on the main face view
✨ Face Style
Live Preview
◉
Classic
Enhanced 2D canvas face
⚙
Robot
3D metallic robot head
👤
Human
Realistic 3D avatar
🎭
VRM
VTuber avatar (VRoid)
Robot and Human styles require WebGL. Changes apply on next page load.
🤖 AI Model
Checking...
Local runs on GPU server (zero API cost), Cloud requires API keys
Running on GPU server via vLLM (4x L40S)
🧠 Context & Memory
Automatically gather context, intent, and setting information
Retrieve relevant past conversations from pgvector
Number of relevant memories to retrieve (1-10)
🔌 API Status
Dialog Service (Chat) :8093
Vision Service (Face/Emotion) :8085
Perception (Whisper STT) :8086
TTS Service (Speech) :8087
Memory (Indexer/pgvector) :8011
Context Service (Weather/Location) :8097
Learner (Content Ingestion) :8016
Observer (Passive Learning) :8017
Summarizer (Memory Consolidation) :8012
Dream Service (Nightly Learning) :8096 🌙 Open Dashboard
Emotional State (CESS) :8102
Relations / Auth :8092
Calendar Service :8105
Privacy & Data