Back to search
aru-voice-call
MCP serverGitHub
Real-time voice call for AI companions — talk to your AI like a phone call. Supports emotion detection, real-time transcription, and ElevenLabs TTS. Works as MCP server for Aru integration.
Safety scan
A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.
Checked today
Badge for your README
Markdown
[](/listings/brittanyyy77--aru-voice-call)HTML
<a href="/listings/brittanyyy77--aru-voice-call"><img src="/api/listings/brittanyyy77--aru-voice-call/badge.svg" alt="findsafeskills safety rating" /></a>Install
i
Install instructions from the README
As written by the repo's author. findsafeskills has not verified these steps.
1. Clone and configure
git clone https://github.com/brittanyyy77/aru-voice-call.git cd aru-voice-call cp .env.example .env # Edit .env with your API keys
2. Run with Docker
docker compose up -d
3. Open in browser
Go to `https://your-vps-ip:8443` on your phone.
What was scanned
The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.
README.md · README · 119 lines
- # aru-voice-call
- Real-time voice call for AI companions — talk to your AI like a phone call.
- ## Features
- - 🎙️ Real-time voice transcription (cloud or local)
- - 🎭 Emotion detection from voice (pitch, speed, volume, pauses)
- - 🔊 AI voice reply via ElevenLabs or MiniMax TTS
- - 📱 Works on mobile browsers, no app needed
- - 🐳 One-command Docker deployment
- - 🌏 China-friendly: supports Aliyun, Volcengine, Baidu STT + MiniMax TTS
- - 🔌 Optional MCP server for Aru integration
- ## How It Works
- The entire call loop runs autonomously on your VPS:
- ```
- You speak → Browser captures audio
- → WebSocket streams to VPS
- → STT transcribes (cloud API or local whisper)
- → librosa analyzes emotion
- → LLM generates reply (with emotion context)
- → TTS synthesizes voice (ElevenLabs or MiniMax)
- → Audio streams back to your browser
- → You hear the reply
- ```
- **No MCP needed for the call itself.** Fully real-time, no middleware delay.
- ## Quick Start
- ### 1. Clone and configure
- ```bash
- git clone https://github.com/brittanyyy77/aru-voice-call.git
- cd aru-voice-call
- cp .env.example .env
- # Edit .env with your API keys
- ```
- ### 2. Run with Docker
- ```bash
- docker compose up -d
- ```
- ### 3. Open in browser
- Go to `https://your-vps-ip:8443` on your phone.
- ## Providers
- ### Speech-to-Text (STT)
- | Provider | Region | Free Tier | Config |
- |----------|--------|-----------|--------|
- | Groq | International | Yes | `STT_PROVIDER=groq` |
- | Deepgram | International | Yes | `STT_PROVIDER=deepgram` |
- | Aliyun | China | Pay-as-you-go | `STT_PROVIDER=aliyun` + `STT_APPKEY` |
- | Volcengine | China | Free quota | `STT_PROVIDER=volcengine` + `STT_APPID` |
- | Baidu | China | Free quota | `STT_PROVIDER=baidu` |
- | Local Whisper | Any | Free (CPU/GPU) | `STT_MODE=local` |
- ### Text-to-Speech (TTS)
- | Provider | Region | Config |
- |----------|--------|--------|
- | ElevenLabs | International | `TTS_PROVIDER=elevenlabs` |
- | MiniMax | China | `TTS_PROVIDER=minimax` |
- ## Configuration
- See `.env.example` for all options. Key settings:
- ```env
- # International setup
- STT_PROVIDER=groq
- TTS_PROVIDER=elevenlabs
- # China setup (no foreign card needed)
- STT_PROVIDER=volcengine
- TTS_PROVIDER=minimax
- ```
- ### SSL Certificate
- Mobile browsers require HTTPS for microphone access:
- ```bash
- mkdir certs
- # Copy your fullchain.pem and privkey.pem into certs/
- ```
- ## Optional: Aru MCP Integration
- If you use [Aru](https://aru.app), bind this as an MCP server:
- - URL: `https://your-vps:8443/mcp`
- - Transport: `streamable-http`
- MCP tools:
- - `get_call_status` — Check active/completed calls
- - `get_call_transcript` — Get full transcript with emotions after call
- - `push_message` — Say something into an active call
- - `get_live_emotion` — Check user's mood during call
- ## Tech Stack
- - **Backend**: Python, FastAPI, WebSocket
- - **STT**: Groq / Deepgram / Aliyun / Volcengine / Baidu / local whisper
- - **Emotion**: librosa audio feature extraction
- - **TTS**: ElevenLabs / MiniMax streaming
- - **Frontend**: Vanilla HTML/JS, Web Audio API
- - **Deploy**: Docker Compose
- ## License
- MIT
Ratings and comments
Sign in to rate this skill.
Sign inFree to use. If it’s helped, pass it on.