Configuration Guide
Configure Beep.AI.Server for your environment
Configuration Files
Beep.AI.Server uses a layered configuration system:
| File | Purpose | Priority |
|---|---|---|
.env | Environment variables (secrets, overrides) | Highest |
config/app_config.json | Persistent configuration (managed by UI) | Medium |
app/config_manager.py | Default values and schema | Lowest |
Environment Variables
| Variable | Description | Default |
|---|---|---|
SQLALCHEMY_DATABASE_URI | Database connection string | sqlite:///app.db |
SECRET_KEY | Flask session secret key | (generated) |
BEEP_JWT_SECRET | JWT token signing key | (generated) |
BEEP_OAUTH2_CLIENT_SECRET | OAuth2 client credentials | — |
BEEP_LDAP_BIND_PASSWORD | LDAP bind password | — |
BEEP_AZURE_AD_CLIENT_SECRET | Azure AD client secret | — |
BEEP_APP_DIR | Override app directory (Docker) | (auto-detected) |
BEEP_SYNC_APP_SERVICES | Synchronous bootstrap | false |
CELERY_ENABLED | Enable Celery task queue | false |
CELERY_BROKER_URL | Celery broker URL | redis://localhost:6379/0 |
REDIS_MANAGED | Auto-start Redis in-app | false |
REDIS_HOST | Redis host | localhost |
REDIS_PORT | Redis port | 6379 |
ASYNC_MODE | SocketIO async mode | threading |
PYTEST_CURRENT_TEST | Set automatically during tests | — |
Database Configuration
SQLite (Default)
SQLALCHEMY_DATABASE_URI=sqlite:///app.dbPostgreSQL
SQLALCHEMY_DATABASE_URI=postgresql://user:password@localhost:5432/beep_aiSQL Server
SQLALCHEMY_DATABASE_URI=mssql+pyodbc://user:password@localhost/beep_ai?driver=ODBC+Driver+17+for+SQL+ServerMySQL
SQLALCHEMY_DATABASE_URI=mysql+pymysql://user:password@localhost:3306/beep_aiAuth Configuration
OAuth2 / OIDC
BEEP_OAUTH2_CLIENT_ID=your-client-id
BEEP_OAUTH2_CLIENT_SECRET=your-client-secret
BEEP_OAUTH2_PROVIDER_URL=https://provider.example.comLDAP
BEEP_LDAP_SERVER=ldap://ldap.example.com
BEEP_LDAP_BIND_DN=cn=admin,dc=example,dc=com
BEEP_LDAP_BIND_PASSWORD=your-passwordAzure AD
BEEP_AZURE_AD_TENANT_ID=your-tenant-id
BEEP_AZURE_AD_CLIENT_ID=your-client-id
BEEP_AZURE_AD_CLIENT_SECRET=your-client-secretEnvironment Isolation
Each AI service uses its own Python virtual environment, managed automatically through the admin UI:
- LLM —
llm_requirements.txt,llm_env_config.json - RAG —
rag_requirements.txt - Agents —
agents_packages.json,agents_env_config.json - MCP —
mcp_packages.json,mcp_env_config.json - Audio —
audio_packages.json - Document Extraction —
doc_extraction_packages.json - Vision —
vision_env_config.json - Text-to-Image —
text_to_image_env_config.json - Tooling —
tooling_packages.json,tooling_env_config.json - Studio —
studio_packages.json
Note
Environments are created on-demand through the admin UI. Each service's overview page shows environment status and handles package installation.