diff --git a/tooling/pipeline/runpod/Dockerfile b/tooling/pipeline/runpod/Dockerfile
index 2842dc3..09fa455 100644
--- a/tooling/pipeline/runpod/Dockerfile
+++ b/tooling/pipeline/runpod/Dockerfile
@@ -22,7 +22,7 @@ RUN curl -L https://github.com/ggml-org/llama.cpp/archive/refs/tags/b9012.tar.gz
 # Pull llama.cpp binaries to use during build if needed
 COPY --from=ghcr.io/ggml-org/llama.cpp:full-cuda /app/lib*.so* /usr/local/lib/
 
-WORKDIR /workspace/app
+WORKDIR /app
 COPY . .
 
 # Build the C++ pipeline
@@ -34,6 +34,8 @@ FROM nvidia/cuda:12.6.3-runtime-ubuntu24.04 AS runtime
 
 # Install only necessary runtime shared libraries
 RUN apt-get update && apt-get install -y --no-install-recommends \
+  curl \
+  ca-certificates \
   libboost-json1.83.0 \
   libboost-program-options1.83.0 \
   libgomp1 \
@@ -41,20 +43,20 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
   zlib1g \
   && rm -rf /var/lib/apt/lists/*
 
-ENV APP_ROOT=/workspace/app \
+ENV APP_ROOT=/app \
   LD_LIBRARY_PATH="/usr/local/lib:${LD_LIBRARY_PATH}"
 
-WORKDIR /workspace/app/build
+WORKDIR /app/build
 
 # Copy only the compiled binaries from the builder
-COPY --from=builder /workspace/app/build/biergarten-pipeline ./
+COPY --from=builder /app/build/biergarten-pipeline ./
 
 # Copy required config files
-COPY locations.json /workspace/app/build/
-COPY beer-styles.json /workspace/app/build/
+COPY locations.json /app/build/
+COPY beer-styles.json /app/build/
 
 # Copy prompt templates
-COPY prompts /workspace/app/prompts
+COPY prompts /app/prompts
 
 # Copy only the necessary shared libraries from builder/llama-bin
 COPY --from=ghcr.io/ggml-org/llama.cpp:full-cuda /app/lib*.so* /usr/local/lib/
diff --git a/tooling/pipeline/runpod/start.sh b/tooling/pipeline/runpod/start.sh
index 1ae4f6c..5a59d8e 100644
--- a/tooling/pipeline/runpod/start.sh
+++ b/tooling/pipeline/runpod/start.sh
@@ -5,8 +5,8 @@ set -e
 MODEL_PATH="${BIERGARTEN_MODEL_PATH:-/workspace/models/google_gemma-4-E4B-it-Q6_K.gguf}"
 OUTPUT_DIR="${BIERGARTEN_OUTPUT_DIR:-/workspace/output}"
 LOG_PATH="${BIERGARTEN_LOG_PATH:-/workspace/logs/pipeline.log}"
-EXECUTABLE="/workspace/app/build/biergarten-pipeline"
-PROMPT_DIR="/workspace/app/prompts"
+EXECUTABLE="/app/build/biergarten-pipeline"
+PROMPT_DIR="/app/prompts"
 
 echo "--- Starting Biergarten Pipeline Environment Check ---"
 
@@ -14,14 +14,6 @@ echo "--- Starting Biergarten Pipeline Environment Check ---"
 mkdir -p "$OUTPUT_DIR"
 mkdir -p "$(dirname "$LOG_PATH")"
 
-# 2. Check for model file
-if [ ! -f "$MODEL_PATH" ]; then
-    echo "ERROR: Model not found at $MODEL_PATH"
-    echo "Current /workspace/models contents:"
-    ls -lh /workspace/models 2>/dev/null || echo "(directory does not exist)"
-    exit 1
-fi
-
 # 3. Build the command arguments
 ARGS=(
     "--model"      "$MODEL_PATH"
@@ -44,6 +36,24 @@ if [[ -n "$BIERGARTEN_EXTRA_ARGS" ]]; then
 fi
 
 echo "--- Executing: $EXECUTABLE ${ARGS[*]} ---"
+# 2. Ensure model directory exists
+mkdir -p "$(dirname "$MODEL_PATH")"
 
+# 3. Download model if missing
+if [ ! -f "$MODEL_PATH" ]; then
+    echo "Model not found. Downloading..."
+
+    curl -L \
+      -o "$MODEL_PATH" \
+      "https://huggingface.co/bartowski/google_gemma-4-E4B-it-GGUF/resolve/main/google_gemma-4-E4B-it-Q6_K.gguf?download=true"
+
+    echo "Download complete."
+fi
+
+# 4. Verify model exists
+if [ ! -f "$MODEL_PATH" ]; then
+    echo "ERROR: Model still not found after download attempt."
+    exit 1
+fi
 # Execute the binary directly, replacing the shell process
 exec "$EXECUTABLE" "${ARGS[@]}"