ad-studio ghcr.io/chrisv006/ad-studio:latest bridge sh false https://github.com/Chrisv006/ad-studio/issues https://github.com/Chrisv006/ad-studio [b]AD Studio[/b] adds audio description (narration of what is on screen) to videos. It watches the video with a vision model (Ollama or Gemini), writes short descriptions, voices them with Kokoro and mixes them into a descriptive audio track (or a longer video that pauses for them). Nothing is kept on the server: results are downloaded and deleted when the browser tab closes. Needs Whisper, Ollama and Kokoro services reachable over HTTP. MediaApp:Video Tools: Productivity: http://[IP]:[PORT:8585]/ https://raw.githubusercontent.com/Chrisv006/ad-studio/main/unraid/ad-studio.xml https://raw.githubusercontent.com/Chrisv006/ad-studio/main/ad_studio/assets/icon.png --user 99:100 8585 /mnt/user/appdata/ad-studio /mnt/cache/appdata/ad-studio-scratch local qwen3-vl:8b-instruct gemini-2.5-flash user 5 4 8 on pauses 1.0 20 20 2.5 8 2 6 af_heart 1.0 10 6