diff --git a/comfyui-audio/Dockerfile b/comfyui-audio/Dockerfile
new file mode 100644
index 0000000..68187cb
--- /dev/null
+++ b/comfyui-audio/Dockerfile
@@ -0,0 +1,44 @@
+FROM nvidia/cuda:12.4.1-cudnn-runtime-ubuntu22.04
+
+ENV DEBIAN_FRONTEND=noninteractive
+ENV PYTHONUNBUFFERED=1
+
+# System dependencies
+RUN apt-get update && apt-get install -y \
+    python3.10 \
+    python3-pip \
+    python3.10-venv \
+    git \
+    wget \
+    ffmpeg \
+    libsndfile1 \
+    && rm -rf /var/lib/apt/lists/*
+
+RUN ln -s /usr/bin/python3.10 /usr/bin/python
+
+WORKDIR /app
+
+ENV GIT_TERMINAL_PROMPT=0
+
+# Clone ComfyUI
+RUN git clone https://github.com/comfyanonymous/ComfyUI.git
+
+WORKDIR /app/ComfyUI
+
+# Install ComfyUI dependencies
+RUN pip install --no-cache-dir "torch>=2.3.0" torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Install ComfyUI-Manager (für einfache Node-Installation über die UI)
+RUN git clone https://github.com/ltdrdata/ComfyUI-Manager.git custom_nodes/ComfyUI-Manager
+RUN pip install --no-cache-dir -r custom_nodes/ComfyUI-Manager/requirements.txt
+
+# Build tools für PyAV (audiocraft dependency)
+RUN apt-get update && apt-get install -y pkg-config libavformat-dev libavcodec-dev libavdevice-dev libavutil-dev libswscale-dev libswresample-dev libavfilter-dev && rm -rf /var/lib/apt/lists/*
+
+# AudioCraft (MusicGen, AudioGen von Meta)
+RUN pip install --no-cache-dir audiocraft
+
+EXPOSE 8188
+
+CMD ["python", "main.py", "--listen", "0.0.0.0", "--port", "8188"]
diff --git a/comfyui-audio/Dockerfile.amd b/comfyui-audio/Dockerfile.amd
new file mode 100644
index 0000000..fcea31f
--- /dev/null
+++ b/comfyui-audio/Dockerfile.amd
@@ -0,0 +1,35 @@
+FROM rocm/pytorch:rocm6.2_ubuntu22.04_py3.10_pytorch_release_2.3.0
+
+ENV DEBIAN_FRONTEND=noninteractive
+ENV PYTHONUNBUFFERED=1
+ENV GIT_TERMINAL_PROMPT=0
+ENV HSA_OVERRIDE_GFX_VERSION=11.0.0
+
+# System dependencies
+RUN apt-get update && apt-get install -y \
+    git wget ffmpeg \
+    libsndfile1 \
+    pkg-config libavformat-dev libavcodec-dev libavdevice-dev \
+    libavutil-dev libswscale-dev libswresample-dev libavfilter-dev \
+    && rm -rf /var/lib/apt/lists/*
+
+WORKDIR /app
+
+# Clone ComfyUI
+RUN git clone https://github.com/comfyanonymous/ComfyUI.git
+
+WORKDIR /app/ComfyUI
+
+# ComfyUI dependencies (torch already included in base image)
+RUN pip install --no-cache-dir -r requirements.txt
+
+# ComfyUI-Manager
+RUN git clone https://github.com/ltdrdata/ComfyUI-Manager.git custom_nodes/ComfyUI-Manager
+RUN pip install --no-cache-dir -r custom_nodes/ComfyUI-Manager/requirements.txt
+
+# AudioCraft
+RUN pip install --no-cache-dir audiocraft
+
+EXPOSE 8188
+
+CMD ["python", "main.py", "--listen", "0.0.0.0", "--port", "8188"]
diff --git a/comfyui-audio/docker-compose.amd.yml b/comfyui-audio/docker-compose.amd.yml
new file mode 100644
index 0000000..0d09798
--- /dev/null
+++ b/comfyui-audio/docker-compose.amd.yml
@@ -0,0 +1,21 @@
+services:
+  comfyui:
+    build:
+      context: .
+      dockerfile: Dockerfile.amd
+    container_name: comfyui-audio
+    ports:
+      - "8188:8188"
+    volumes:
+      - ./models:/app/ComfyUI/models
+      - ./output:/app/ComfyUI/output
+      - ./input:/app/ComfyUI/input
+      - ./custom_nodes:/app/ComfyUI/custom_nodes
+    devices:
+      - /dev/kfd
+      - /dev/dri
+    group_add:
+      - video
+    environment:
+      - HSA_OVERRIDE_GFX_VERSION=11.0.0
+    restart: unless-stopped
diff --git a/comfyui-audio/docker-compose.yml b/comfyui-audio/docker-compose.yml
new file mode 100644
index 0000000..12f6b63
--- /dev/null
+++ b/comfyui-audio/docker-compose.yml
@@ -0,0 +1,21 @@
+services:
+  comfyui:
+    build: .
+    container_name: comfyui-audio
+    ports:
+      - "8188:8188"
+    volumes:
+      - ./models:/app/ComfyUI/models
+      - ./output:/app/ComfyUI/output
+      - ./input:/app/ComfyUI/input
+      - ./custom_nodes:/app/ComfyUI/custom_nodes
+    environment:
+      - NVIDIA_VISIBLE_DEVICES=all
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              count: all
+              capabilities: [gpu]
+    restart: unless-stopped
diff --git a/comfyui-audio/setup.md b/comfyui-audio/setup.md
new file mode 100644
index 0000000..d4d9016
--- /dev/null
+++ b/comfyui-audio/setup.md
@@ -0,0 +1,34 @@
+# ComfyUI + AudioCraft Setup
+
+## Voraussetzungen
+- Docker Desktop mit WSL2 Backend
+- NVIDIA Container Toolkit (in Docker Desktop integriert ab Version 4.x)
+
+## Einmalig: WSL2 NVIDIA Support aktivieren
+In Docker Desktop → Settings → Resources → GPU → Enable GPU support aktivieren.
+
+## Starten
+```bash
+cd comfyui-audio
+docker compose up --build
+```
+Beim ersten Start lädt es alle Abhängigkeiten (~5-10 min).
+
+## ComfyUI öffnen
+http://localhost:8188
+
+## AudioCraft Modelle laden
+Die Modelle werden automatisch beim ersten Nutzen heruntergeladen und unter `./models` gecacht.
+
+Empfohlene Modelle für Game Audio:
+- `musicgen-medium` — Musik/Atmosphäre (1.5GB)
+- `musicgen-stereo-medium` — Stereo Musik
+- `audiogen-medium` — Sound Effects (Schritte, Umgebungsgeräusche)
+
+## Workflow für das Game
+1. Atmosphärische Musik: musicgen mit Prompt wie *"dark dungeon ambience, slow, mysterious"*
+2. Kampfmusik: *"epic battle music, drums, dark fantasy"*
+3. Sound Effects via audiogen: *"sword clash metal", "footsteps stone floor"*
+
+## Output
+Generierte Audiodateien landen unter `./output/`