Optimize: use --prefer-binary for wheels on Linux but not committed

Dockerfile

@@ -1,18 +1,18 @@
-FROM ghcr.io/abetlen/llama-cpp-python:latest
+FROM python:3.11-slim
 
 WORKDIR /app
 
-# Install additional system dependencies
+# Install system dependencies (minimal, no build tools for base)
 RUN apt-get update && apt-get install -y \
     tesseract-ocr \
     libtesseract-dev \
     && rm -rf /var/lib/apt/lists/*
 
-# Copy requirements (llama-cpp-python NOT included - already in base image!)
+# Copy requirements
 COPY requirements.txt .
 
-# Install remaining dependencies
-RUN pip install --no-cache-dir -r requirements.txt
+# Install with binary preference (uses wheels if available, falls back to source)
+RUN pip install --no-cache-dir --prefer-binary -r requirements.txt
 
 # Copy app
 COPY . .

requirements.txt

@@ -1,5 +1,6 @@
 fastapi
 uvicorn
+llama-cpp-python
 supabase
 python-multipart
 pytesseract