audio-cpp ghcr.io/0xshug0/audio.cpp:full-cpu https://github.com/0xShug0/audio.cpp/pkgs/container/audio.cpp bridge bash false true AI Tools: http://[IP]:[PORT:8080]/ https://raw.githubusercontent.com/lozenge0/audio-cpp-unraid/main/assets/icon.svg https://github.com/0xShug0/audio.cpp https://github.com/lozenge0/audio-cpp-unraid/issues https://raw.githubusercontent.com/lozenge0/audio-cpp-unraid/main/templates/audio-cpp.xml https://raw.githubusercontent.com/lozenge0/audio-cpp-unraid/main/README.md audio.cpp for Unraid: community-maintained installation templates using unmodified upstream Docker images and the native WebUI/audio API. Choose CPU, NVIDIA CUDA 12, or NVIDIA CUDA 13 when installing. No models or personal configuration are bundled. Images follow upstream rolling Docker builds, not exclusively numbered releases. BETA: isolated runtime and private installation tests passed on one host; public branch selection and full deployment acceptance remain pending. See the README for tested images and limitations. Trusted-network access only: this setup provides no authentication. Runs as UID 99/GID 100 using Docker --user; model storage must be writable by this identity. Model downloads require network access and adequate storage. CPU variant: no GPU runtime required. Review the README before installation. CPU — no GPU required; upstream rolling image --user=99:100 server --ui --ui-management --host 0.0.0.0 --port 8080 --backend cpu 8080 /mnt/user/appdata/audio-cpp/models full-cuda12 NVIDIA — CUDA 12; upstream rolling image ghcr.io/0xshug0/audio.cpp:full-cuda12 NVIDIA Driver plugin with a compatible GPU and host driver, and the NVIDIA container runtime. Trusted-network access only: no authentication. Runs as UID 99/GID 100 using Docker --user; model storage must be writable by this identity. Hardware and model compatibility require validation; see README. --user=99:100 --runtime=nvidia server --ui --ui-management --host 0.0.0.0 --port 8080 --backend cuda 8080 /mnt/user/appdata/audio-cpp/models all compute,utility full-cuda13 NVIDIA — CUDA 13; compatible GPU and newer driver required ghcr.io/0xshug0/audio.cpp:full-cuda13 NVIDIA Driver plugin, a CUDA 13-compatible GPU and host driver, and the NVIDIA container runtime. CUDA 13 excludes pre-Turing architectures; newer hardware alone does not establish compatibility. Trusted-network access only: no authentication. Runs as UID 99/GID 100 using Docker --user; model storage must be writable by this identity. See README. --user=99:100 --runtime=nvidia server --ui --ui-management --host 0.0.0.0 --port 8080 --backend cuda 8080 /mnt/user/appdata/audio-cpp/models all compute,utility