feat: add portable backend manager and static file server with hardware telemetry support

README.md

@@ -1,6 +1,6 @@
 # 🖼️ Local AI Image Generator
 
-### An easy, zero-setup Stable Diffusion GUI for Windows and Linux. Run GGUF & Safetensors models offline without Python configuration.
+### An easy, zero-setup Stable Diffusion GUI for Windows, Linux, and macOS. Run GGUF & Safetensors models offline without Python configuration.
 
 
 
@@ -22,7 +22,7 @@
 ---
 
 ## 📖 Overview
-**Local AI Image Generator** is a zero-configuration, portable desktop environment for running Stable Diffusion (Safetensors/GGUF/CKPT) offline on Windows and Linux. Double-clicking `start.bat` (Windows) or running `./start.sh` (Linux) automatically handles dependency setup, GPU backend matching, and launches a high-performance local web workspace.
+**Local AI Image Generator** is a zero-configuration, portable desktop environment for running Stable Diffusion (Safetensors/GGUF/CKPT) offline on Windows, Linux, and macOS. Double-clicking `start.bat` (Windows) or running `./start.sh` (Linux/macOS) automatically handles dependency setup, GPU backend matching, and launches a high-performance local web workspace.
 
 ---
 
@@ -40,11 +40,17 @@
 3. **Add Models:** Drop `.safetensors`, `.gguf`, or `.ckpt` weights into `app/models/` (or download them via the **Model Manager** tab in the UI).
 4. **Generate:** Open `http://localhost:1420` in your browser, select your model, and write a prompt.
 
+### macOS
+1. **Check compatibility:** The prebuilt macOS backend is for **Apple Silicon (M1 or newer)** and uses **Metal** GPU acceleration.
+2. **Launch:** Open Terminal in the project folder and run **`./start.sh`** (downloads portable Node.js and the pre-compiled Metal backend on first run).
+3. **Add Models:** Drop `.safetensors`, `.gguf`, or `.ckpt` weights into `app/models/` (or download them via the **Model Manager** tab in the UI).
+4. **Generate:** Open `http://localhost:1420` in your browser, select your model, and write a prompt.
+
 ---
 
 ## ✨ Features
 *   **100% Offline & Private:** Inference runs completely locally on your hardware.
-*   **Auto-Detected GPU Acceleration:** Configures **CUDA** for Nvidia cards, **ROCm** for AMD cards, and **Vulkan** for AMD/Intel/NVIDIA fallback.
+*   **Auto-Detected GPU Acceleration:** Configures **CUDA** for Nvidia cards, **ROCm** for AMD cards, **Vulkan** for AMD/Intel/NVIDIA fallback, and **Metal** for Apple Silicon Macs.
 *   **Zero System Footprint:** Node.js is sandboxed inside the folder. No global environment paths are altered.
 *   **Integrated Model Manager:** Paste a Hugging Face URL to download weights directly, or drag-and-drop local weight files to import them.
 *   **Real-time Telemetry:** Monitor RAM, VRAM, CPU, and GPU load directly in the UI.
@@ -56,16 +62,16 @@
 ```
 local-ai-image-generator/
 ├── start.bat                  # Main double-click entrypoint (Windows)
-├── start.sh                   # Main terminal entrypoint (Linux)
+├── start.sh                   # Main terminal entrypoint (Linux/macOS)
 ├── PLAN.md                    # Linux port implementation plan
 ├── LICENSE                    # MIT Open Source license
 ├── .gitignore
 ├── README.md                  
 ├── scripts/
 │   ├── setup.ps1              # Automated GPU-detect and environment installer (Windows)
-│   ├── setup.sh               # Automated GPU-detect and environment installer (Linux)
+│   ├── setup.sh               # Automated GPU-detect and environment installer (Linux/macOS)
 │   ├── reset.ps1              # Cleans runtime environments (Windows)
-│   ├── reset.sh               # Cleans runtime environments (Linux)
+│   ├── reset.sh               # Cleans runtime environments (Linux/macOS)
 │   └── serve.cjs              # UI web server and backend lifecycle manager
 └── app/
     ├── frontend/              # UI source code (Vite + React)
@@ -95,9 +101,17 @@ local-ai-image-generator/
 | **Intel Arc / integrated** | Vulkan | CPU | Cross-vendor Vulkan support. |
 | **Integrated / None** | CPU | — | Runs on logical CPU threads (slow). |
 
+### macOS
+
+| Hardware | Primary | Fallback | Notes |
+| :--- | :--- | :--- | :--- |
+| **Apple Silicon (M1 or newer)** | Metal | CPU | Uses the official Darwin arm64 stable-diffusion.cpp backend. |
+| **Intel Mac** | Source build required | CPU | Official prebuilt macOS backend is Apple Silicon only. |
+
 **System requirements:**
 - **glibc 2.38 or newer** is required for the prebuilt Linux backends (Ubuntu 24.04, Fedora 40+, etc.).
 - The setup script will warn you if your glibc is older. You can still run setup, but the prebuilt backends will not start.
+- **Apple Silicon (M1 or newer)** is required for the prebuilt macOS Metal backend.
 
 **Linux setup modes:**
 - **Default (`./start.sh`)**: Downloads CPU + Vulkan backends (~120–150 MB).
@@ -118,7 +132,7 @@ Typical generation times for an image with **20 steps** (e.g. 512x512 resolution
 ---
 
 ## 🛠️ Troubleshooting
-*   **Reset Environment:** If a build fails or you want to clear dependencies, run `scripts/reset.ps1` (Windows) or `scripts/reset.sh` (Linux). (This preserves your models and generated images).
+*   **Reset Environment:** If a build fails or you want to clear dependencies, run `scripts/reset.ps1` (Windows) or `scripts/reset.sh` (Linux/macOS). (This preserves your models and generated images).
 *   **Port Conflicts:** The frontend uses `1420` by default. The backend tries `8080` first, then automatically falls back to a free port if `8080` is already busy.
 *   **Linux backends fail to start with `GLIBC_2.38' not found`:** The prebuilt binaries require glibc 2.38+ (Ubuntu 24.04). Upgrade your distribution or build stable-diffusion.cpp from source (see below).
 *   **Linux ROCm not loading:** Make sure your AMD GPU and kernel are compatible with ROCm 7.13. The app will automatically fall back to Vulkan if ROCm cannot initialize.
@@ -130,11 +144,14 @@ Typical generation times for an image with **20 steps** (e.g. 512x512 resolution
 
 If your distribution has an older glibc than 2.38, or you want a CUDA backend on Linux, you can build `stable-diffusion.cpp` directly on your machine. The resulting binary will be linked against your system's glibc and will not have the compatibility issues of the prebuilt releases.
 
+For macOS, the included `scripts/build_from_source.sh` builds the Metal backend and copies it to `app/backend/mac/sd`.
+
 ### Requirements
 - `git`, `cmake`, `make` (or `ninja`), and a C++17 compiler (`g++` / `clang++`).
 - For **CUDA**: the NVIDIA CUDA toolkit (`nvcc`) must be on your `PATH`.
 - For **Vulkan**: the Vulkan SDK / loader and a compatible driver.
 - For **ROCm**: AMD ROCm development libraries.
+- For **macOS Metal**: Apple Command Line Tools or Xcode.
 
 ### Build commands
 
@@ -157,8 +174,11 @@ cmake .. -DSD_VULKAN=ON -DSD_BUILD_SHARED_LIBS=ON -DCMAKE_BUILD_TYPE=Release
 # ROCm
 cmake .. -DSD_HIPBLAS=ON -DSD_BUILD_SHARED_LIBS=ON -DCMAKE_BUILD_TYPE=Release
 
+# macOS Metal
+cmake .. -DSD_METAL=ON -DSD_BUILD_SHARED_LIBS=ON -DCMAKE_BUILD_TYPE=Release
+
 # 3. Build
-cmake --build . --config Release -j$(nproc)
+cmake --build . --config Release -j$(getconf _NPROCESSORS_ONLN 2>/dev/null || sysctl -n hw.ncpu)
 
 # 4. Copy the binaries into this project
 cp bin/sd* /path/to/Local-AI-Image-Generator/app/backend/linux/<backend>/

app/dist/index.html

@@ -1,18 +1,18 @@
-<!doctype html>
-<html lang="en">
-  <head>
-    <meta charset="UTF-8" />
-    <link rel="icon" type="image/svg+xml" href="/vite.svg" />
-    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
-    <link rel="preconnect" href="https://fonts.googleapis.com">
-    <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
-    <link href="https://fonts.googleapis.com/css2?family=Outfit:wght@300;400;500;600;700&display=swap" rel="stylesheet">
-    <title>Local AI Image Generator</title>
+<!doctype html>
+<html lang="en">
+  <head>
+    <meta charset="UTF-8" />
+    <link rel="icon" type="image/svg+xml" href="/vite.svg" />
+    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
+    <link rel="preconnect" href="https://fonts.googleapis.com">
+    <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
+    <link href="https://fonts.googleapis.com/css2?family=Outfit:wght@300;400;500;600;700&display=swap" rel="stylesheet">
+    <title>Local AI Image Generator</title>
     <script type="module" crossorigin src="/assets/index-CG3WZqfX.js"></script>
     <link rel="stylesheet" crossorigin href="/assets/index-DW6G7wc2.css">
-  </head>
-
-  <body>
-    <div id="root"></div>
-  </body>
-</html>
+  </head>
+
+  <body>
+    <div id="root"></div>␍
+  </body>
+</html>

app/frontend/package-lock.json

@@ -1554,9 +1554,9 @@
       }
     },
     "node_modules/electron-to-chromium": {
-      "version": "1.5.369",
-      "resolved": "https://registry.npmjs.org/electron-to-chromium/-/electron-to-chromium-1.5.369.tgz",
-      "integrity": "sha512-XM22K9FNaaCOvMMrBn1caIc8v0g6+pKt660ZbfQqUZvfil0hEzr8ZoiY7VcSLGM3L/x3rz5PqZrk+bKOOmVM9w==",
+      "version": "1.5.370",
+      "resolved": "https://registry.npmjs.org/electron-to-chromium/-/electron-to-chromium-1.5.370.tgz",
+      "integrity": "sha512-D5tSHJReAb/Kf3Hu9F/GO4lJuSWzEWHwvQ/kKSUP7pimNgvxkSKj+gUQhHpKKACwrin7rS3byU7IxreF56rl5g==",
       "dev": true,
       "license": "ISC"
     },

scripts/build_from_source.sh

@@ -1,7 +1,7 @@
 #!/usr/bin/env bash
 #
 # Local AI Image Generator - Compile backends from source
-# For systems with GLIBC < 2.38 (e.g. Ubuntu 22.04)
+# For Linux systems with GLIBC < 2.38 or macOS systems that need a local Metal build.
 #
 
 set -euo pipefail
@@ -9,8 +9,14 @@ set -euo pipefail
 SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 ROOT_DIR="$(dirname "$SCRIPT_DIR")"
 APP_DIR="$ROOT_DIR/app"
-BACKEND_DIR="$APP_DIR/backend/linux"
+PLATFORM="$(uname -s)"
+if [ "$PLATFORM" = "Darwin" ]; then
+    BACKEND_DIR="$APP_DIR/backend/mac"
+else
+    BACKEND_DIR="$APP_DIR/backend/linux"
+fi
 BUILD_DIR="$APP_DIR/tools/build-sd"
+JOBS="$(getconf _NPROCESSORS_ONLN 2>/dev/null || sysctl -n hw.ncpu 2>/dev/null || echo 4)"
 
 # Pinned release tag
 PINNED_TAG="master-685-19bdfe2"
@@ -35,12 +41,36 @@ git fetch origin
 git checkout -f "$PINNED_TAG"
 git submodule update --init --recursive
 
+if [ "$PLATFORM" = "Darwin" ]; then
+    echo ""
+    echo "=== Building macOS Metal Backend ==="
+    rm -rf build-metal && mkdir build-metal && cd build-metal
+    cmake .. -DSD_METAL=ON -DSD_BUILD_SHARED_LIBS=ON -DCMAKE_BUILD_TYPE=Release
+    cmake --build . --config Release -j"$JOBS"
+
+    echo "Copying macOS Metal backend..."
+    mkdir -p "$BACKEND_DIR"
+    if [ -f bin/sd-server ]; then
+        cp bin/sd-server "$BACKEND_DIR/sd"
+    elif [ -f bin/sd ]; then
+        cp bin/sd "$BACKEND_DIR/sd"
+    else
+        echo "ERROR: bin/sd-server or bin/sd was not found!"
+        exit 1
+    fi
+    find . -name "*.dylib" -exec cp {} "$BACKEND_DIR/" \; 2>/dev/null || true
+    chmod +x "$BACKEND_DIR/sd"
+    echo ""
+    echo "=== macOS Metal build completed successfully! ==="
+    exit 0
+fi
+
 # 2. Build CPU Backend
 echo ""
 echo "=== Building CPU Backend ==="
 rm -rf build-cpu && mkdir build-cpu && cd build-cpu
 cmake .. -DSD_BUILD_SHARED_LIBS=ON -DCMAKE_BUILD_TYPE=Release
-cmake --build . --config Release -j$(nproc)
+cmake --build . --config Release -j"$JOBS"
 
 echo "Copying CPU binaries..."
 mkdir -p "$BACKEND_DIR/cpu"
@@ -71,7 +101,7 @@ echo ""
 echo "=== Building Vulkan Backend ==="
 if rm -rf build-vulkan && mkdir build-vulkan && cd build-vulkan && \
    cmake .. -DSD_VULKAN=ON -DSD_BUILD_SHARED_LIBS=ON -DCMAKE_BUILD_TYPE=Release && \
-   cmake --build . --config Release -j$(nproc); then
+   cmake --build . --config Release -j"$JOBS"; then
 
     echo "Copying Vulkan binaries..."
     mkdir -p "$BACKEND_DIR/vulkan"
@@ -102,7 +132,7 @@ echo ""
 echo "=== Building CUDA Backend ==="
 rm -rf build-cuda && mkdir build-cuda && cd build-cuda
 cmake .. -DSD_CUDA=ON -DSD_BUILD_SHARED_LIBS=ON -DCMAKE_BUILD_TYPE=Release
-cmake --build . --config Release -j$(nproc)
+cmake --build . --config Release -j"$JOBS"
 
 echo "Copying CUDA binaries..."
 mkdir -p "$BACKEND_DIR/cuda"

scripts/reset.sh

@@ -1,6 +1,6 @@
 #!/usr/bin/env bash
 #
-# Local AI Image Generator - Linux Reset Script
+# Local AI Image Generator - Linux/macOS Reset Script
 # Resets portable app dependencies/builds while preserving user models and outputs.
 #
 

scripts/serve.cjs

@@ -28,10 +28,10 @@ const BACKEND_PATHS = {
   cuda: path.join(ROOT, "app", "backend", "win", "cuda", "sd-cuda.exe"),
   vulkan: path.join(ROOT, "app", "backend", "win", "vulkan", "sd-vulkan.exe"),
   mac: path.join(ROOT, "app", "backend", "mac", "sd"),
-  linuxCpu: path.join(ROOT, "app", "backend", "linux", "cpu", "sd-cpu"),
-  linuxVulkan: path.join(ROOT, "app", "backend", "linux", "vulkan", "sd-vulkan"),
-  linuxRocm: path.join(ROOT, "app", "backend", "linux", "rocm", "sd-rocm"),
-  linuxCuda: path.join(ROOT, "app", "backend", "linux", "cuda", "sd-cuda"),
+  linuxCpu: path.join(ROOT, "app", "backend", "linux", "cpu", "sd-server-cpu"),
+  linuxVulkan: path.join(ROOT, "app", "backend", "linux", "vulkan", "sd-server-vulkan"),
+  linuxRocm: path.join(ROOT, "app", "backend", "linux", "rocm", "sd-server-rocm"),
+  linuxCuda: path.join(ROOT, "app", "backend", "linux", "cuda", "sd-server-cuda"),
 };
 let BACKEND_PATH = "";
 if (osPlatform === "win32") {
@@ -188,6 +188,21 @@ function getGpuInfo() {
     }
   }
 
+  if (osPlatform === "darwin") {
+    try {
+      const output = execSync(
+        "system_profiler SPDisplaysDataType | awk -F': ' '/Chipset Model|Vendor/ {print $2; exit}'",
+        { encoding: "utf8", stdio: ["ignore", "pipe", "ignore"] },
+      ).trim();
+      if (output) {
+        cachedGpuInfo = { name: output };
+        return cachedGpuInfo;
+      }
+    } catch (_) {}
+    cachedGpuInfo = { name: "Apple Metal GPU" };
+    return cachedGpuInfo;
+  }
+
   cachedGpuInfo = { name: "Unavailable" };
   return cachedGpuInfo;
 }
@@ -588,8 +603,11 @@ function getBackendOptions() {
   const rocmInstalled = osPlatform === "linux" && fs.existsSync(BACKEND_PATHS.linuxRocm);
   const rocmAvailable = rocmInstalled && hasAmdGpu() && backendAccepts(BACKEND_PATHS.linuxRocm, "rocm");
   const cpuInstalled = osPlatform === "linux" && fs.existsSync(BACKEND_PATHS.linuxCpu);
+  const metalInstalled = osPlatform === "darwin" && fs.existsSync(BACKEND_PATHS.mac);
+  const metalAvailable = metalInstalled && backendAccepts(BACKEND_PATHS.mac, "metal");
 
   const options = [{ id: "cpu", label: "CPU", available: true }];
+  if (metalAvailable) options.push({ id: "metal", label: "Metal GPU", available: true });
   if (vulkanAvailable) options.push({ id: "vulkan", label: "Vulkan GPU", available: true });
   if (rocmAvailable) options.push({ id: "rocm", label: "ROCm GPU (AMD)", available: true });
   if (cudaAvailable) options.push({ id: "cuda", label: "CUDA GPU", available: true });
@@ -604,9 +622,14 @@ function getBackendOptions() {
   if (rocmInstalled && !rocmAvailable) {
     unavailable.push({ id: "rocm", label: "ROCm GPU (AMD)", reason: "Installed, but ROCm backend validation failed." });
   }
+  if (metalInstalled && !metalAvailable) {
+    unavailable.push({ id: "metal", label: "Metal GPU", reason: "Installed, but Metal backend validation failed." });
+  }
 
   let defaultBackend = "cpu";
-  if (cudaAvailable) {
+  if (metalAvailable) {
+    defaultBackend = "metal";
+  } else if (cudaAvailable) {
     const gpuName = String(getGpuInfo().name).toLowerCase();
     const isGtxCard = gpuName.includes("gtx");
     if (isGtxCard && vulkanAvailable) {
@@ -626,6 +649,7 @@ function getBackendOptions() {
     cudaAvailable,
     vulkanAvailable,
     rocmAvailable,
+    metalAvailable,
     defaultBackendType: defaultBackend,
   };
   return cachedBackendOptions;
@@ -640,7 +664,17 @@ function backendAccepts(binaryPath, backendName) {
       "--model", path.join(MODELS, "__backend_probe_missing__.safetensors"),
       "--listen-port", "18082",
     ];
-    let result = spawnSync(binaryPath, probeArgs, { encoding: "utf8", timeout: 5000 });
+    const spawnEnv = { ...process.env };
+    if (osPlatform === "linux") {
+      const dir = path.dirname(binaryPath);
+      const existing = spawnEnv.LD_LIBRARY_PATH || "";
+      spawnEnv.LD_LIBRARY_PATH = dir + (existing ? ":" + existing : "");
+    } else if (osPlatform === "darwin") {
+      const dir = path.dirname(binaryPath);
+      const existing = spawnEnv.DYLD_LIBRARY_PATH || "";
+      spawnEnv.DYLD_LIBRARY_PATH = dir + (existing ? ":" + existing : "");
+    }
+    let result = spawnSync(binaryPath, probeArgs, { encoding: "utf8", timeout: 5000, env: spawnEnv });
     let output = `${result.stdout || ""}\n${result.stderr || ""}`;
 
     // Some binaries do not support --backend. If we see "unknown argument",
@@ -650,7 +684,7 @@ function backendAccepts(binaryPath, backendName) {
         "--model", path.join(MODELS, "__backend_probe_missing__.safetensors"),
         "--listen-port", "18082",
       ];
-      result = spawnSync(binaryPath, fallbackArgs, { encoding: "utf8", timeout: 5000 });
+      result = spawnSync(binaryPath, fallbackArgs, { encoding: "utf8", timeout: 5000, env: spawnEnv });
       output = `${result.stdout || ""}\n${result.stderr || ""}`;
     }
 
@@ -691,7 +725,11 @@ function selectBackendPath(useGpu, backendType = "auto") {
     if (fs.existsSync(BACKEND_PATHS.linuxCpu)) return BACKEND_PATHS.linuxCpu;
     return BACKEND_PATH;
   }
-  // macOS fallback
+  if (osPlatform === "darwin") {
+    if (fs.existsSync(BACKEND_PATHS.mac)) return BACKEND_PATHS.mac;
+    return BACKEND_PATH;
+  }
+  // generic fallback
   if (fs.existsSync(BACKEND_PATHS.mac)) return BACKEND_PATHS.mac;
   return BACKEND_PATH;
 }
@@ -709,6 +747,7 @@ function getBackendMode(backendPath, useGpu, backendType = "auto") {
   if (name.includes("cuda")) return "CUDA GPU";
   if (name.includes("rocm")) return "ROCm GPU";
   if (name.includes("vulkan")) return "Vulkan GPU";
+  if (osPlatform === "darwin" || backendType === "metal") return "Metal GPU";
   return "GPU";
 }
 
@@ -882,6 +921,13 @@ async function startBackend(settings = {}) {
       "--rng", "cpu",
       "--sampler-rng", "cpu",
     );
+  } else if (requestedBackend === "metal") {
+    args.push(
+      "--backend", "metal",
+      "--params-backend", "metal",
+      "--rng", "cpu",
+      "--sampler-rng", "cpu",
+    );
   }
 
   if (currentSettings.vaeTiling) {
@@ -909,6 +955,10 @@ async function startBackend(settings = {}) {
       const existing = spawnEnv.LD_LIBRARY_PATH || "";
       spawnEnv.LD_LIBRARY_PATH = extraLibs.join(":") + (existing ? ":" + existing : "");
     }
+  } else if (osPlatform === "darwin") {
+    const existing = spawnEnv.DYLD_LIBRARY_PATH || "";
+    const backendDir = path.dirname(BACKEND_PATHS.mac);
+    spawnEnv.DYLD_LIBRARY_PATH = backendDir + (existing ? ":" + existing : "");
   }
 
   console.log("  [backend] Starting:", path.basename(backendPath), args.join(" "));
@@ -1825,7 +1875,7 @@ const server = http.createServer(async (req, res) => {
   });
 });
 
-server.listen(PORT_FRONTEND, "127.0.0.1", () => {
+server.listen(PORT_FRONTEND, "0.0.0.0", () => {
   console.log("");
   console.log("  ============================================================");
   console.log("   LOCAL AI IMAGE GENERATOR  |  Running");

scripts/setup.sh

@@ -1,6 +1,6 @@
 #!/usr/bin/env bash
 #
-# Local AI Image Generator - Linux Setup Script
+# Local AI Image Generator - Linux/macOS Setup Script
 # Self-contained: no apt/yum/pacman, no global Node.js install.
 #
 
@@ -11,18 +11,40 @@ ROOT_DIR="$(dirname "$SCRIPT_DIR")"
 APP_DIR="$ROOT_DIR/app"
 FRONTEND_DIR="$APP_DIR/frontend"
 TOOLS_DIR="$APP_DIR/tools"
-NODE_DIR="$TOOLS_DIR/node-linux"
+DIST_DIR="$APP_DIR/dist"
+PLATFORM="$(uname -s)"
+ARCH="$(uname -m)"
+
+if [[ "$PLATFORM" == "Darwin" ]]; then
+  PLATFORM_LABEL="macOS"
+  NODE_DIR="$TOOLS_DIR/node-mac"
+  BACKEND_DIR="$APP_DIR/backend/mac"
+else
+  PLATFORM_LABEL="Linux"
+  NODE_DIR="$TOOLS_DIR/node-linux"
+  BACKEND_DIR="$APP_DIR/backend/linux"
+fi
+
 NODE_BIN="$NODE_DIR/bin/node"
 NPM_BIN="$NODE_DIR/bin/npm"
-DIST_DIR="$APP_DIR/dist"
-BACKEND_DIR="$APP_DIR/backend/linux"
 
 # Release pins
 SD_RELEASE="master-685-19bdfe2"
 SD_SHORT_HASH="${SD_RELEASE##*-}"
 SD_BASE_URL="https://github.com/leejet/stable-diffusion.cpp/releases/download/$SD_RELEASE"
 NODE_VERSION="22.12.0"
-NODE_TARBALL="node-v${NODE_VERSION}-linux-x64.tar.xz"
+
+if [[ "$PLATFORM" == "Darwin" ]]; then
+  if [[ "$ARCH" == "arm64" ]]; then
+    NODE_PLATFORM_ARCH="darwin-arm64"
+  else
+    NODE_PLATFORM_ARCH="darwin-x64"
+  fi
+  NODE_TARBALL="node-v${NODE_VERSION}-${NODE_PLATFORM_ARCH}.tar.gz"
+else
+  NODE_PLATFORM_ARCH="linux-x64"
+  NODE_TARBALL="node-v${NODE_VERSION}-${NODE_PLATFORM_ARCH}.tar.xz"
+fi
 NODE_URL="https://nodejs.org/dist/v${NODE_VERSION}/$NODE_TARBALL"
 
 # Flags
@@ -36,7 +58,7 @@ print_header() {
   clear 2>/dev/null || true
   echo ""
   echo "  ============================================================"
-  echo "   LOCAL AI IMAGE GENERATOR  -  Linux First-Time Setup"
+  echo "   LOCAL AI IMAGE GENERATOR  -  $PLATFORM_LABEL First-Time Setup"
   echo "   100% Self-Contained  |  No System Install Required"
   echo "  ============================================================"
   echo ""
@@ -180,15 +202,39 @@ copy_binaries_from_extracted() {
   find "$extracted_dir" -type f -name "*.so" -exec cp {} "$dest_dir/" \; 2>/dev/null || true
 }
 
+copy_macos_backend_from_extracted() {
+  local extracted_dir="$1" dest_dir="$2"
+  local target="$dest_dir/sd"
+  local server_bin=""
+  local cli_bin=""
+
+  server_bin="$(find "$extracted_dir" -type f -name "sd-server" | head -n 1)"
+  cli_bin="$(find "$extracted_dir" -type f \( -name "sd" -o -name "sd-cli" \) | head -n 1)"
+
+  if [[ -n "$server_bin" ]]; then
+    cp "$server_bin" "$target"
+  elif [[ -n "$cli_bin" ]]; then
+    cp "$cli_bin" "$target"
+  else
+    print_fail "No sd-server, sd, or sd-cli binary was found in the macOS backend archive."
+    return 1
+  fi
+
+  find "$extracted_dir" -type f \( -name "*.dylib" -o -name "*.metallib" \) -exec cp {} "$dest_dir/" \; 2>/dev/null || true
+  chmod +x "$target" 2>/dev/null || true
+}
+
 # ════════════════════════════════════════════════════════════════════════════
 print_header
 
-check_glibc
+if [[ "$PLATFORM" == "Linux" ]]; then
+  check_glibc
+fi
 
 TOTAL_STEPS=4
 
 # ── Step 1: Portable Node.js ────────────────────────────────────────────────
-print_step 1 $TOTAL_STEPS "Setting up portable Node.js (app/tools/node-linux/)"
+print_step 1 $TOTAL_STEPS "Setting up portable Node.js ($NODE_DIR/)"
 
 if [[ -x "$NODE_BIN" && -x "$NPM_BIN" ]]; then
   VERSION=$("$NODE_BIN" --version)
@@ -201,7 +247,7 @@ else
   extract_tarxz "$NODE_TAR_PATH" "$TOOLS_DIR" "Node.js"
   rm -f "$NODE_TAR_PATH"
 
-  EXTRACTED_DIR="$(find "$TOOLS_DIR" -maxdepth 1 -type d -name 'node-v*-linux-x64' | head -n 1)"
+  EXTRACTED_DIR="$(find "$TOOLS_DIR" -maxdepth 1 -type d -name "node-v*-${NODE_PLATFORM_ARCH}" | head -n 1)"
   if [[ -d "$EXTRACTED_DIR" ]]; then
     rm -rf "$NODE_DIR"
     mv "$EXTRACTED_DIR" "$NODE_DIR"
@@ -217,11 +263,32 @@ else
 fi
 
 # ── Step 2: stable-diffusion.cpp Backends ───────────────────────────────────
-VENDOR="$(detect_gpu_vendor)"
-print_step 2 $TOTAL_STEPS "Detecting GPU vendor: ${VENDOR:-none}"
-
 mkdir -p "$BACKEND_DIR"
 
+if [[ "$PLATFORM" == "Darwin" ]]; then
+  print_step 2 $TOTAL_STEPS "Setting up stable-diffusion.cpp Metal backend (app/backend/mac/)"
+  if [[ "$ARCH" != "arm64" ]]; then
+    print_fail "The official macOS backend binary is Apple Silicon only (arm64)."
+    print_info "Intel Macs need a local source build with Metal/OpenBLAS and a matching app/backend/mac/sd binary."
+    exit 1
+  fi
+
+  MAC_BACKEND="$BACKEND_DIR/sd"
+  if [[ -x "$MAC_BACKEND" ]]; then
+    print_ok "macOS Metal backend already ready."
+  else
+    MAC_ZIP="$TOOLS_DIR/sd-mac-metal.zip"
+    download_file "$SD_BASE_URL/sd-master-${SD_SHORT_HASH}-bin-Darwin-macOS-15.7.7-arm64.zip" "$MAC_ZIP" "stable-diffusion.cpp Metal Backend (macOS arm64)"
+    extract_zip "$MAC_ZIP" "$BACKEND_DIR/extracted" "macOS Metal Backend"
+    rm -f "$MAC_ZIP"
+    copy_macos_backend_from_extracted "$BACKEND_DIR/extracted" "$BACKEND_DIR"
+    rm -rf "$BACKEND_DIR/extracted"
+    print_ok "macOS Metal backend installed."
+  fi
+else
+  VENDOR="$(detect_gpu_vendor)"
+  print_step 2 $TOTAL_STEPS "Detecting GPU vendor: ${VENDOR:-none}"
+
 # CPU backend (always)
 CPU_BACKEND_DIR="$BACKEND_DIR/cpu"
 if [[ ! -f "$CPU_BACKEND_DIR/sd-cpu" ]]; then
@@ -284,6 +351,7 @@ if [[ $MAX_PERF -eq 1 ]] && [[ "$VENDOR" == "nvidia" ]]; then
   print_info "Linux CUDA backend is disabled pending a reliable upstream binary."
   print_info "NVIDIA GPUs on Linux will use the Vulkan backend instead."
 fi
+fi
 
 # ── Step 3: npm install ─────────────────────────────────────────────────────
 print_step 3 $TOTAL_STEPS "Installing frontend dependencies (app/frontend/)"
@@ -320,7 +388,7 @@ echo "   Setup complete! Run ./start.sh to launch."
 echo "  ============================================================"
 echo ""
 
-if [[ $MAX_PERF -eq 0 ]] && [[ "$VENDOR" == "nvidia" || "$VENDOR" == "amd" ]]; then
+if [[ "$PLATFORM" == "Linux" && $MAX_PERF -eq 0 ]] && [[ "$VENDOR" == "nvidia" || "$VENDOR" == "amd" ]]; then
   echo "  Tip: For maximum GPU performance, re-run with:"
   echo "       ./scripts/setup.sh --max-perf"
   echo ""

start.sh

@@ -1,15 +1,25 @@
 #!/usr/bin/env bash
 #
-# Local AI Image Generator - Linux Launcher
+# Local AI Image Generator - Linux/macOS Launcher
 # Double-click or run: ./start.sh
-# Use --max-perf to enable ROCm/CUDA backend downloads on first setup.
+# Use --max-perf to enable ROCm backend downloads on Linux first setup.
 #
 
 set -euo pipefail
 
 SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 APP_DIR="$SCRIPT_DIR/app"
-NODE_DIR="$APP_DIR/tools/node-linux"
+PLATFORM="$(uname -s)"
+if [[ "$PLATFORM" == "Darwin" ]]; then
+  NODE_DIR="$APP_DIR/tools/node-mac"
+  BACKEND_PATH="$APP_DIR/backend/mac/sd"
+  PLATFORM_LABEL="macOS"
+else
+  NODE_DIR="$APP_DIR/tools/node-linux"
+  BACKEND_PATH="$APP_DIR/backend/linux/vulkan/sd-vulkan"
+  CPU_BACKEND_PATH="$APP_DIR/backend/linux/cpu/sd-cpu"
+  PLATFORM_LABEL="Linux"
+fi
 NODE_BIN="$NODE_DIR/bin/node"
 DIST_INDEX="$APP_DIR/dist/index.html"
 SETUP_SCRIPT="$SCRIPT_DIR/scripts/setup.sh"
@@ -38,15 +48,21 @@ if [[ ! -f "$DIST_INDEX" ]]; then
   SETUP_REASON="Frontend build is missing."
 fi
 
-# At minimum we need CPU or Vulkan backend on Linux
-if [[ ! -f "$APP_DIR/backend/linux/cpu/sd-cpu" && ! -f "$APP_DIR/backend/linux/vulkan/sd-vulkan" ]]; then
-  SETUP_REASON="No Linux backend binary is installed."
+if [[ "$PLATFORM" == "Darwin" ]]; then
+  if [[ ! -x "$BACKEND_PATH" ]]; then
+    SETUP_REASON="No macOS Metal backend binary is installed."
+  fi
+else
+  # At minimum we need CPU or Vulkan backend on Linux
+  if [[ ! -f "$CPU_BACKEND_PATH" && ! -f "$BACKEND_PATH" ]]; then
+    SETUP_REASON="No Linux backend binary is installed."
+  fi
 fi
 
 if [[ -n "$SETUP_REASON" ]]; then
   echo ""
   echo "  ============================================================"
-  echo "   LOCAL AI IMAGE GENERATOR  |  $SETUP_MODE"
+  echo "   LOCAL AI IMAGE GENERATOR  |  $PLATFORM_LABEL $SETUP_MODE"
   echo "  ============================================================"
   echo ""
   if [[ "$SETUP_MODE" == "First-Time Setup" ]]; then
@@ -98,7 +114,10 @@ SERVER_PID=$!
 sleep 2
 
 # Open browser
-if command -v xdg-open >/dev/null 2>&1; then
+if [[ "$PLATFORM" == "Darwin" ]] && command -v open >/dev/null 2>&1; then
+  echo "  Opening browser at http://localhost:${FRONTEND_PORT}"
+  open "http://localhost:${FRONTEND_PORT}" >/dev/null 2>&1 &
+elif command -v xdg-open >/dev/null 2>&1; then
   echo "  Opening browser at http://localhost:${FRONTEND_PORT}"
   xdg-open "http://localhost:${FRONTEND_PORT}" >/dev/null 2>&1 &
 else