backend/.env.example

# ============================================================
# Document Processing AI Agent — backend configuration
# Copy to `.env` and fill in. Only GEMINI_API_KEY is required.
# ============================================================

# --- LLM providers ---
# Default provider is the free Gemini tier. Get a key at https://aistudio.google.com/apikey
GEMINI_API_KEY=
# Optional switchable fallback. Leave blank to disable.
OPENAI_API_KEY=

# Which provider to use by default: "gemini" or "openai"
DEFAULT_LLM_PROVIDER=gemini

# Model names (override only if you know what you want)
GEMINI_MODEL=gemini-2.0-flash
OPENAI_MODEL=gpt-4o-mini

# Embedding model for RAG (Gemini, free)
GEMINI_EMBED_MODEL=text-embedding-004

# --- Storage ---
DATA_DIR=./data
CHROMA_DIR=./data/chroma

# --- Server ---
HOST=0.0.0.0
PORT=8000
# Comma-separated allowed origins for CORS (frontend dev server)
CORS_ORIGINS=http://localhost:3000

# --- Extraction ---
# Max pages to render to images for the viewer / vision fallback
MAX_RENDER_PAGES=30
# Run Docling OCR on scanned PDFs/images (slower but handles scans)
ENABLE_OCR=true