mediasage/docker-compose.yml at main · atomthkun/mediasage · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
services:
  mediasage:
    image: ecwilson/mediasage:latest
    # Alternative registries:
    # image: ghcr.io/ecwilsonaz/mediasage:latest
    # To build locally instead, comment out 'image' and uncomment 'build':
    # build: .
    container_name: mediasage
    # Optional: Run as specific user for NAS/Synology (use your PUID:PGID)
    # user: "1000:1000"
    ports:
      - "5765:5765"
    environment:
      # Required: Plex server settings
      - PLEX_URL=${PLEX_URL}
      - PLEX_TOKEN=${PLEX_TOKEN}
      - PLEX_MUSIC_LIBRARY=${PLEX_MUSIC_LIBRARY:-Music}

      # Performance tuning
      - UVICORN_WORKERS=${UVICORN_WORKERS:-1}

      # LLM Provider (anthropic, openai, gemini, ollama, or custom)
      # Defaults to gemini. Can also be configured in the web UI.
      #- LLM_PROVIDER=gemini

      # Cloud provider API keys (set one)
      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY:-}
      - OPENAI_API_KEY=${OPENAI_API_KEY:-}
      - GEMINI_API_KEY=${GEMINI_API_KEY:-}

      # Optional: Override default models
      #- LLM_MODEL_ANALYSIS=gemini-2.5-flash
      #- LLM_MODEL_GENERATION=gemini-2.5-flash

      # Local LLM: Ollama (experimental)
      #- LLM_PROVIDER=ollama
      #- OLLAMA_URL=http://host.docker.internal:11434
      #- OLLAMA_CONTEXT_WINDOW=32768

      # Local LLM: Custom OpenAI-compatible API (experimental)
      #- LLM_PROVIDER=custom
      #- CUSTOM_LLM_URL=https://openrouter.ai/api/v1
      #- CUSTOM_LLM_API_KEY=sk-or-your-key
      #- CUSTOM_CONTEXT_WINDOW=32768
    volumes:
      # Persist library cache across restarts
      # Ensure directory is writable: either set 'user' above, or run:
      #   mkdir -p ./data && chmod 777 ./data
      - ./data:/app/data
      # Optional: persist UI settings
      # - ./config.user.yaml:/app/config.user.yaml
    restart: unless-stopped