Jotist-NVIDIA ghcr.io/jaysqvl/jotist:latest-cuda https://github.com/jaysqvl/Jotist/pkgs/container/jotist bridge bash false --runtime=nvidia https://github.com/jaysqvl/Jotist/issues https://github.com/jaysqvl/Jotist https://raw.githubusercontent.com/jaysqvl/unraid-templates/main/docs/apps/jotist.md Jotist turns audio and video into searchable transcripts, speaker labels, notes, and optional summaries and chat. Run local speech models on your own server and keep your recordings under your control. This NVIDIA option follows the latest stable CUDA 12.6 release. Install Unraid's NVIDIA Driver plugin and use a compatible GPU and host driver. Select all GPUs or enter a GPU UUID in NVIDIA_VISIBLE_DEVICES. RTX 50-series/Blackwell inference is not qualified for this image. Model support and VRAM requirements vary by model. Application data, managed runtimes, and home/model caches are stored in three persistent appdata folders. Open the WebUI after installation to create the first account. Optional cloud transcription and LLM providers send content to the provider you configure. Productivity: Tools:Utilities http://[IP]:[PORT:8080]/ https://raw.githubusercontent.com/jaysqvl/unraid-templates/main/templates/jotist-nvidia.xml https://raw.githubusercontent.com/jaysqvl/unraid-templates/main/icons/jotist.png https://raw.githubusercontent.com/jaysqvl/Jotist/main/web/project-site/public/screenshots/jotist-library-dark.png speech transcription speech-to-text whisper parakeet canary diarization audio stt ai cuda gpu nvidia 64-bit Intel or AMD processor (linux/amd64), CUDA 12.6-compatible NVIDIA GPU and host driver, and the NVIDIA Driver plugin. RTX 50-series/Blackwell inference is not qualified. Internet access and disk space for model downloads. RAM and VRAM requirements vary by model. MIT 8080 /mnt/user/appdata/jotist-nvidia/data /mnt/user/appdata/jotist-nvidia/whisperx-env /mnt/user/appdata/jotist-nvidia/home all compute,utility 99 100 /app/data/jotist.db production