docker-compose.yml

services:
  gateway:
    build:
      context: .
      dockerfile: Dockerfile
    image: multimind-sdk:latest
    ports:
      - "8000:8000"
    env_file:
      - path: .env
        required: false
    environment:
      # Provider API keys — all optional; configure only the ones you use
      - OPENAI_API_KEY=${OPENAI_API_KEY:-}
      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY:-}
      - GROQ_API_KEY=${GROQ_API_KEY:-}
      - MISTRAL_API_KEY=${MISTRAL_API_KEY:-}
      - GEMINI_API_KEY=${GEMINI_API_KEY:-}
      - DEEPSEEK_API_KEY=${DEEPSEEK_API_KEY:-}
      - HUGGINGFACE_API_KEY=${HUGGINGFACE_API_KEY:-}
      # Local models via the optional ollama service (profile: local)
      - OLLAMA_API_BASE=${OLLAMA_API_BASE:-http://ollama:11434}
      - OLLAMA_MODEL_NAME=${OLLAMA_MODEL_NAME:-mistral}
      - DEFAULT_MODEL=${DEFAULT_MODEL:-openai}
      - LOG_LEVEL=${LOG_LEVEL:-INFO}
    volumes:
      - multimind_sessions:/app
    healthcheck:
      test: ["CMD", "python", "-c", "import urllib.request; urllib.request.urlopen('http://127.0.0.1:8000/health', timeout=4)"]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    restart: unless-stopped

  # Optional: local models, no API keys needed.
  # Start with: docker compose --profile local up
  ollama:
    image: ollama/ollama:latest
    profiles: ["local"]
    ports:
      - "11434:11434"
    volumes:
      - ollama_models:/root/.ollama
    healthcheck:
      test: ["CMD", "ollama", "list"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 20s
    restart: unless-stopped

  # Optional: Chroma vector store for RAG workloads.
  # Start with: docker compose --profile vector-store up
  chroma:
    image: chromadb/chroma:latest
    profiles: ["vector-store"]
    ports:
      - "8001:8000"
    volumes:
      - chroma_data:/chroma/chroma
    environment:
      - ANONYMIZED_TELEMETRY=false
    restart: unless-stopped

volumes:
  multimind_sessions:
  ollama_models:
  chroma_data: