Quick Start
This page walks you through a first run of N.E.K.O. after completing the Development Setup.
1. Start the servers
bash
# In separate terminals:
uv run python memory_server.py
uv run python main_server.py2. Configure an API provider
Navigate to http://localhost:48911/api_key and configure at least the Core API provider.
For a quick test without an API key, select Free as the Core API provider.
3. Interact with the default character
Open http://localhost:48911 in your browser. The default character ("小天") will be loaded with a Live2D model.
Text mode: Type a message in the chat input and press Enter.
Voice mode: Click the microphone button to start a voice session. Speak naturally — the system uses server-side VAD (Voice Activity Detection) to detect when you finish speaking.
4. Customize the character
Navigate to http://localhost:48911/chara_manager to:
- Change the character's name, gender, age, and personality traits
- Set a custom Live2D or VRM model
- Clone a custom voice (upload a ~15-second clean audio sample)
- Edit the system prompt for full control over behavior
5. Explore the Web UI pages
| URL | Purpose |
|---|---|
/ | Main chat interface |
/chara_manager | Character settings |
/api_key | API key configuration |
/model_manager | Live2D/VRM model management |
/live2d_emotion_manager | Emotion-to-animation mapping |
/vrm_emotion_manager | VRM emotion mapping |
/voice_clone | Voice cloning |
/memory_browser | Browse and edit memories |
/steam_workshop_manager | Steam Workshop items |
Next steps
- Project Structure — Understand the codebase layout
- Architecture Overview — How the three servers cooperate
- API Reference — All REST and WebSocket endpoints