Configuration
March 2, 2026 ยท View on GitHub
MoneyPrinter reads configuration from .env (project root).
Use .env.example as your template.
Required
| Variable | Description |
|---|---|
TIKTOK_SESSION_ID | TikTok session cookie (sessionid) used for TTS voice endpoint calls. |
PEXELS_API_KEY | API key used to fetch stock video clips. |
Optional
| Variable | Description | Default |
|---|---|---|
IMAGEMAGICK_BINARY | Absolute path to ImageMagick executable. If empty, auto-detected from PATH. | auto-detect |
OLLAMA_BASE_URL | Ollama server base URL used for model listing and chat generation. | http://localhost:11434 |
OLLAMA_MODEL | Fallback model if frontend does not send a model value. | llama3.1:8b |
ASSEMBLY_AI_API_KEY | If set, subtitles are generated with AssemblyAI; otherwise local subtitle generation is used. | empty |
POSTGRES_DB | Database name for Docker Postgres service. | moneyprinter |
POSTGRES_USER | Database user for Docker Postgres service. | moneyprinter |
POSTGRES_PASSWORD | Database password for Docker Postgres service. | moneyprinter |
DATABASE_URL | SQLAlchemy DSN used by API and worker (postgresql+psycopg://... or sqlite:///...). | sqlite:///moneyprinter.db |
Notes
- Ollama models shown in the frontend are fetched from backend endpoint
/api/models, which queriesOLLAMA_BASE_URL/api/tags. - Pull models before use, for example:
ollama pull llama3.1:8b
- If ImageMagick is not discovered automatically, set
IMAGEMAGICK_BINARYexplicitly. - New architecture uses a database-backed job queue. In Docker, use Postgres via
DATABASE_URL.