System Requirements
Before using CueMate, please confirm your device meets the following system requirements.
1. macOS Platform
1.1 Desktop Client Requirements
NOTE
CueMate supports both Intel and Apple Silicon chips on Mac. Please select the corresponding architecture installation package when downloading.
1.1.1 Minimum Requirements
- Operating System: macOS 13.0 (Ventura) or higher
- Processor: Intel Core i5 dual-core or Apple Silicon (M1/M2/M3)
- Memory: 4GB RAM
- Storage: 10GB available space (application itself is about 5GB)
- Display: 1280 x 720 resolution
TIP
Recommended configuration provides a smoother experience, especially when using speech recognition and AI generation features.
1.1.2 Recommended Requirements
- Operating System: macOS 15 (Sequoia) or higher
- Processor: Intel Core i7 quad-core or Apple Silicon (M1 Pro/M2 Pro/M3 Pro)
- Memory: 8GB RAM
- Storage: 15GB available space
- Display: 1920 x 1080 resolution or higher
1.1.3 System Permissions
IMPORTANT
Required Permissions (core features will not work without these):
Microphone Permission - For capturing your voice input
- Grant at: System Preferences > Security & Privacy > Privacy > Microphone
Screen Recording Permission - For capturing interview software system audio
- Grant at: System Preferences > Security & Privacy > Privacy > Screen Recording
Optional Permissions (enhance user experience):
Accessibility Permission - For global shortcuts and floating window features
- Grant at: System Preferences > Security & Privacy > Privacy > Accessibility
Full Disk Access Permission - For log file read/write
- Grant at: System Preferences > Security & Privacy > Privacy > Full Disk Access
1.2 Docker Desktop Requirements
CueMate's backend services run in Docker containers and require Docker Desktop installation.
IMPORTANT
Installation Package Includes Docker Desktop
The CueMate installation package includes Docker Desktop installer (automatically selects Intel or Apple Silicon version based on chip type). The installation wizard will automatically detect and guide you through Docker installation, no manual download required.
1.2.1 Docker Desktop Configuration Requirements
- Docker Desktop Version: 4.0 or higher
- Memory Allocation: At least 4GB (8GB recommended)
- CPU Cores: At least 2 cores (4 cores recommended)
- Storage: 10GB available space (for Docker images and data)
1.2.2 Automated Installation Process
- Open CueMate installation package DMG file
- Run "CueMate Installer.app" installation wizard
- Installation wizard automatically detects Docker Desktop status
- If not installed, wizard guides you to install the bundled Docker Desktop
- After installation, wizard automatically configures resource allocation and starts all services
1.3 Docker Service Resource Usage
CueMate includes 6 Docker containers:
| No. | Service | Minimum Memory | Recommended Memory | CPU Usage |
|---|---|---|---|---|
| 1 | cuemate-web | 128MB | 256MB | 2% |
| 2 | cuemate-web-api | 256MB | 512MB | 5% |
| 3 | cuemate-llm-router | 256MB | 512MB | 5% |
| 4 | cuemate-rag-service | 256MB | 512MB | 5% |
| 5 | cuemate-asr | 1GB | 2GB | 20% |
| 6 | cuemate-chroma | 256MB | 512MB | 5% |
| Total | 2.13GB | 4.25GB | 42% |
WARNING
Resource Usage Notes:
- ASR (speech recognition) service is the most resource-intensive component (1-2GB memory + 20% CPU)
- If your device has limited performance, consider disabling this service and only using LLM Q&A features
- Total resource requirements: Recommend at least 4GB system memory (2GB system + 2GB Docker)
1.4 Network Requirements
1.4.1 Bandwidth Requirements
- Minimum Bandwidth: 5 Mbps (for LLM API calls)
- Recommended Bandwidth: 10 Mbps or higher
1.4.2 Network Connection
- Local Services: No internet connection required (ASR, database, Web API)
- LLM Services: Internet connection required (calling cloud LLM APIs)
- Offline Mode: Supports local models (Ollama, vLLM, Xinference), no internet required
1.4.3 Network Ports
NOTE
CueMate uses the following local ports. Please ensure ports are not occupied by other programs.
| No. | Port | Service | Purpose |
|---|---|---|---|
| 1 | 3001 | Web API | Business API service |
| 2 | 3002 | LLM Router | LLM routing service |
| 3 | 3003 | RAG Service | Knowledge base retrieval service |
| 4 | 3004 | Main Window Service | Admin interface |
| 5 | 8000 | ChromaDB | Vector database |
| 6 | 10095 | cuemate-asr | Speech recognition service |
TIP
Port Conflict Resolution: If a port is occupied, you can modify the port mapping in the Docker Compose configuration file ~/Library/Application\ Support/cuemate-desktop-client/docker/docker-compose.yml.
1.5 Audio Device Requirements
1.5.1 Microphone
- Type: Any standard microphone (built-in or external)
- Sample Rate: 16000 Hz or higher
- Bit Depth: 16 bit
- Channels: Mono or stereo
Recommended Devices:
- Laptop built-in microphone (basic usability)
- USB microphone (better audio quality)
- Headset microphone (good noise reduction)
- Professional microphone (best quality)
1.5.2 System Audio Capture
macOS uses AudioTee or virtual audio devices for system audio capture.
TIP
System Audio Test: After installation, please use the "System Audio Test" feature to verify that you can capture audio from interview software properly.
1.6 Browser Requirements (Main Window Application)
1.6.1 Supported Browsers
- Chrome: Version 90 or higher (recommended)
- Edge: Version 90 or higher
- Safari: Version 14 or higher
- Firefox: Version 88 or higher
1.6.2 Browser Feature Requirements
- WebSocket support
- LocalStorage support
- JavaScript enabled
- Cookies enabled
1.7 External Service Requirements
1.7.1 LLM API (one required)
IMPORTANT
You need at least one LLM API key to use CueMate's AI features.
| No. | Provider | Description |
|---|---|---|
| 1 | OpenAI | International provider of GPT series models |
| 2 | Azure OpenAI | OpenAI service on Microsoft Azure platform |
| 3 | Anthropic | International provider of Claude series models |
| 4 | Google Gemini | Google's multimodal large language model |
| 5 | AWS Bedrock | AWS large language model service platform |
| 6 | Alibaba Cloud Bailian | Alibaba Cloud's enterprise-grade LLM service platform |
| 7 | Qwen | Alibaba Cloud's Qwen series large language model |
| 8 | Zhipu AI | Zhipu AI's GLM series models |
| 9 | Baichuan Intelligence | Baichuan Intelligence's Baichuan series models |
| 10 | Baidu Qianfan | Baidu's large language model platform |
| 11 | ByteDance Doubao | ByteDance's Doubao large model service |
| 12 | iFlytek Spark | iFlytek's Spark cognitive large model |
| 13 | Tencent Hunyuan | Tencent's self-developed large language model |
| 14 | Tencent Cloud Knowledge Engine | Tencent Cloud's LLM service |
| 15 | Kimi | Moonshot AI's Kimi intelligent assistant |
| 16 | MiniMax | MiniMax's ultra-long text large model |
| 17 | DeepSeek | DeepSeek's high-performance large model |
| 18 | SenseTime SenseNova | SenseTime's SenseNova series models |
| 19 | StepFun | Step series models focusing on long context |
| 20 | SiliconFlow | Service platform focused on AI inference acceleration |
| 21 | Regolo | Local model service |
| 22 | Ollama | Open-source tool for running large models locally |
| 23 | vLLM | High-performance large model inference engine |
| 24 | Xinference | Inference framework supporting multiple models |
1.8 Data Storage Space
1.8.1 Minimum Storage Space
- Application: 500MB
- Docker Images: > 5GB
- SQLite Database: 100MB (initial)
- ChromaDB Vector Data: 500MB (initial)
- Log Files: <= 100MB
- Total: <= 10GB
1.8.2 Recommended Storage Space
- Knowledge Base Documents: 1GB (can be increased as needed)
- System Data Growth: 1GB reserved space
- Total: <= 12GB
1.8.3 Data Directory Locations
- Application Data:
~/Library/Application Support/cuemate-desktop-client - SQLite Database:
~/Library/Application Support/cuemate-desktop-client/data/sqlite/cuemate.db - ChromaDB Data: Docker volume
chroma_data - Log Files:
~/Library/Application Support/cuemate-desktop-client/data/logs
1.9 Performance Benchmarks
1.9.1 Desktop Client Performance
Low-end Device (4GB RAM + Dual-core CPU):
- Application startup time: 3-5 seconds
- Speech recognition latency: 1-2 seconds
- AI answer generation latency: 1-3 seconds
- CPU usage: 20-50%
- Memory usage: 2-3GB
High-end Device (8GB RAM + Quad-core CPU):
- Application startup time: 2-3 seconds
- Speech recognition latency: 0.5-1 second
- AI answer generation latency: 1-2 seconds
- CPU usage: 15-30%
- Memory usage: 1.5-2.5GB
1.9.2 Docker Service Performance
Low-end Device:
- Docker startup time: 10-20 seconds
- All services ready time: 20-30 seconds
- Vector retrieval latency: 200-500ms
- Total CPU usage: 40-60%
- Total memory usage: 3-5GB
High-end Device:
- Docker startup time: 5-10 seconds
- All services ready time: 10-20 seconds
- Vector retrieval latency: 50-150ms
- Total CPU usage: 20-40%
- Total memory usage: 4-6GB
1.10 FAQ
1.10.1 Performance Optimization Suggestions
TIP
If device performance is insufficient:
- Close unnecessary background applications
- Reduce Docker container resource allocation
- Use smaller LLM models (e.g., GPT-4o-mini)
- Disable text-to-speech feature
- Reduce the number of documents in vector database
TIP
If network speed is slow:
- Use domestic LLM services (Alibaba Cloud, Zhipu, etc.)
- Lower the LLM max_tokens parameter
- Use local models (Ollama, vLLM)
- Enable LLM response caching
1.10.2 Compatibility Issues
WARNING
macOS version too low:
- If your macOS version is lower than 13, you cannot run CueMate
- Recommend upgrading to the latest macOS version
WARNING
Docker Desktop won't start:
- Check if virtualization support is enabled
- Confirm macOS version meets Docker Desktop requirements
2. Distributed Deployment Server Requirements
NOTE
If you choose Distributed Mode, you need to prepare a Linux server to run Docker services. The desktop client runs on local macOS and connects to the remote server via SSH.
2.1 Server Hardware Requirements
Minimum Configuration:
- CPU: 2 cores
- Memory: 4GB RAM
- Storage: 20GB available space
- Network: 10 Mbps bandwidth
Recommended Configuration:
- CPU: 4 cores
- Memory: 8GB RAM
- Storage: 50GB available space
- Network: 100 Mbps bandwidth
2.2 Server Software Requirements
Operating System:
- Ubuntu: 20.04 LTS or higher (Recommended)
- Debian: 11 or higher
- CentOS: 8 or higher
- Other: Any Linux distribution supporting Docker
Required Software:
- Docker: 20.10 or higher
- Docker Compose: 2.0 or higher
- SSH Server: OpenSSH or compatible SSH service
2.3 Network Configuration
Required Ports:
| Port | Service | Description |
|---|---|---|
| 22 | SSH | SSH connection (configurable) |
| 3001 | Web API | Business API service |
| 3002 | LLM Router | LLM routing service |
| 3003 | RAG Service | Knowledge base retrieval service |
| 3004 | Web | Admin interface |
| 8000 | ChromaDB | Vector database |
| 10095 | ASR | Speech recognition service |
TIP
Service ports (3001-10095) only need to allow access from client IP. SSH port must be accessible from client.
2.4 Deployment Path
- Default Path:
/opt/cuemate - Docker Data:
/opt/cuemate/data - Logs:
/opt/cuemate/logs
3. Windows Platform
Under Development
The Windows version is currently under development, please stay tuned.
If you have any suggestions or requirements for the Windows version, feel free to provide feedback through:
- GitHub Issues: https://github.com/cuemate-chat/cuemate/issues
- Email: nuneatonhydroplane@gmail.com
If your device meets the above requirements, you can start installing CueMate.
For other questions, please check FAQ or contact technical support.