jdevto · jajera · Dec 13, 2025 · Dec 12, 2025
diff --git a/README.md b/README.md
@@ -11,6 +11,7 @@ Command-line interface tools/scripts
 - [install_grafana.sh](docs/install_grafana.md)
 - [install_kafka.sh](docs/install_kafka.md)
 - [install_kubectl.sh](docs/install_kubectl.md)
+- [install_lm_studio.sh](docs/install_lm_studio.md)
 - [install_loki.sh](docs/install_loki.md)
 - [install_promtail.sh](docs/install_promtail.md)
 - [install_python.sh](docs/install_python.md)
@@ -40,6 +41,7 @@ Refer to the respective files for detailed usage instructions:
 - **[install_grafana.sh](docs/install_grafana.md)**: Install or uninstall Grafana web-based analytics and monitoring platform.
 - **[install_kafka.sh](docs/install_kafka.md)**: Install or uninstall Apache Kafka on a Linux system.
 - **[install_kubectl.sh](docs/install_kubectl.md)**: Install or uninstall `kubectl` on a Linux system.
+- **[install_lm_studio.sh](docs/install_lm_studio.md)**: Install or uninstall LM Studio for running large language models locally with headless API server support.
 - **[install_loki.sh](docs/install_loki.md)**: Install or uninstall Grafana Loki log aggregation system.
 - **[install_promtail.sh](docs/install_promtail.md)**: Install or uninstall Grafana Promtail log shipper with flexible configuration options.
 - **[install_python.sh](docs/install_python.md)**: Install or uninstall Python with version pinning support on Linux, macOS, and Windows.

diff --git a/docs/install_lm_studio.md b/docs/install_lm_studio.md
@@ -0,0 +1,311 @@
+# install_lm_studio.sh
+
+This script installs, configures, and manages LM Studio on Linux systems. LM Studio is a desktop application for running large language models (LLMs) locally. This script installs LM Studio in headless mode with an API server for programmatic access.
+
+## Usage
+
+```bash
+./install_lm_studio.sh [install|uninstall|start|stop|status]
+```
+
+- **install** (default): Installs LM Studio AppImage, bootstraps the CLI, and creates a systemd service.
+- **uninstall**: Removes LM Studio from the system.
+- **start**: Starts the LM Studio service.
+- **stop**: Stops the LM Studio service.
+- **status**: Shows the service status.
+
+## Optional Environment Variables
+
+- **LM_STUDIO_VERSION**: Pin specific version (default: `0.3.34-1`)
+- **LM_STUDIO_PORT**: API server port (default: `1234`)
+- **LM_STUDIO_USER**: User to run service as (auto-detected if not set)
+
+## Example Usage
+
+To install LM Studio with default settings:
+
+```bash
+./install_lm_studio.sh install
+```
+
+To install with a custom port:
+
+```bash
+LM_STUDIO_PORT=8080 ./install_lm_studio.sh install
+```
+
+To install a specific version:
+
+```bash
+LM_STUDIO_VERSION=0.3.34-1 ./install_lm_studio.sh install
+```
+
+To install with a specific user:
+
+```bash
+LM_STUDIO_USER=myuser ./install_lm_studio.sh install
+```
+
+To start the service:
+
+```bash
+./install_lm_studio.sh start
+```
+
+To check service status:
+
+```bash
+./install_lm_studio.sh status
+```
+
+To stop the service:
+
+```bash
+./install_lm_studio.sh stop
+```
+
+To uninstall LM Studio:
+
+```bash
+./install_lm_studio.sh uninstall
+```
+
+## Running Without Cloning
+
+```bash
+bash <(curl -s https://raw.githubusercontent.com/jdevto/cli-tools/main/scripts/install_lm_studio.sh) install
+```
+
+## Verification
+
+After installation, check if LM Studio CLI is installed:
+
+```bash
+lms --version
+```
+
+Check the service status:
+
+```bash
+sudo systemctl status lm-studio
+```
+
+The API server will be available at: <http://localhost:1234> (or your configured port)
+
+## Supported Operating Systems
+
+- **Linux** (AppImage format)
+  - Ubuntu (uses `apt` for package management)
+  - Red Hat / Fedora (uses `dnf` for package management)
+  - CentOS / Amazon Linux (uses `yum` for package management)
+
+**Note**: LM Studio AppImage is only available for Linux. macOS and Windows are not supported by this script.
+
+## Supported Architectures
+
+- **x64** (x86_64)
+- **arm64** (aarch64)
+
+## Features
+
+- Automatically detects package manager and operating system
+- Dynamically detects the current user (ec2-user, ubuntu, or current user)
+- Installs required dependencies (fuse, Xvfb, GUI libraries) if missing
+- Downloads LM Studio AppImage (~1GB, may take several minutes)
+- Bootstraps LM Studio CLI by running AppImage in headless mode
+- Creates systemd service for automatic startup
+- Runs AppImage with Xvfb (virtual framebuffer) for headless operation
+- Installs CLI globally at `/usr/local/bin/lms`
+- Automatically starts and enables the service after installation
+- Skips installation if LM Studio is already installed
+- Handles version pinning for reproducible installations
+
+## Error Handling
+
+- If LM Studio is already installed, the script will skip reinstallation and display the current version.
+- If an unsupported OS or architecture is detected, the script exits with an error message.
+- The script validates that the service file and binary exist before starting the service.
+- Missing dependencies are automatically installed.
+- User detection handles various scenarios (root, sudo, regular user).
+- Download failures are handled gracefully with resume support.
+
+## Cleanup
+
+- Temporary installation files are automatically removed after execution.
+- The script uses `trap cleanup EXIT` to ensure cleanup even if interrupted.
+- Temporary directories in `/tmp/lm-studio-install` are cleaned up.
+
+## Configuration
+
+The script automatically configures LM Studio with:
+
+- **Systemd service** at `/etc/systemd/system/lm-studio.service`
+- **Service wrapper script** at `/usr/local/bin/lm-studio-service.sh`
+- **AppImage location** at `/opt/lm-studio/LM-Studio-{VERSION}-{ARCH}.AppImage`
+- **Global CLI binary** at `/usr/local/bin/lms`
+- **User CLI binary** at `~/.lmstudio/bin/lms`
+- **Xvfb virtual display** on `:99` for headless operation
+- **API server** on the configured port (default: 1234)
+
+### Configuration Details
+
+- **Installation directory**: `/opt/lm-studio`
+- **AppImage binary**: `/opt/lm-studio/LM-Studio-{VERSION}-{ARCH}.AppImage`
+- **Symlink**: `/usr/local/bin/lm-studio` (points to AppImage)
+- **CLI binary**: `/usr/local/bin/lms` (global) and `~/.lmstudio/bin/lms` (user)
+- **Service user**: Automatically detected (ec2-user, ubuntu, or current user)
+- **Working directory**: User's home directory
+- **Port**: Configurable via `LM_STUDIO_PORT` environment variable (default: 1234)
+- **Display**: Xvfb virtual display `:99` for headless operation
+
+## Service Management
+
+After installation, you can manage the LM Studio service using:
+
+```bash
+sudo systemctl start lm-studio
+sudo systemctl stop lm-studio
+sudo systemctl restart lm-studio
+sudo systemctl enable lm-studio
+sudo systemctl disable lm-studio
+```
+
+Or use the script commands:
+
+```bash
+./install_lm_studio.sh start
+./install_lm_studio.sh stop
+./install_lm_studio.sh status
+```
+
+### Service Commands
+
+- **Start**: `sudo systemctl start lm-studio` or `./install_lm_studio.sh start`
+- **Stop**: `sudo systemctl stop lm-studio` or `./install_lm_studio.sh stop`
+- **Status**: `sudo systemctl status lm-studio` or `./install_lm_studio.sh status`
+- **Enable**: `sudo systemctl enable lm-studio` (starts automatically on boot)
+- **Disable**: `sudo systemctl disable lm-studio` (prevents auto-start on boot)
+
+## Prerequisites
+
+The script automatically installs required dependencies:
+
+- **wget or curl** - for downloading the AppImage
+- **fuse** - for mounting AppImage files
+- **Xvfb** - virtual framebuffer for headless operation
+- **GUI libraries** - required for AppImage bootstrap (gtk3, atk, nss, etc.)
+- **nodejs and npm** - for CLI functionality
+
+### Dependencies by Package Manager
+
+**dnf/yum (RHEL/Fedora/Amazon Linux):**
+
+- fuse, fuse-libs
+- xorg-x11-server-Xvfb
+- xdg-utils, atk, at-spi2-atk, gtk3, nss, cups-libs
+- libXScrnSaver, gdk-pixbuf2, alsa-lib
+- nodejs, npm
+
+**apt/apt-get (Ubuntu/Debian):**
+
+- libfuse2
+- xvfb
+- xdg-utils, libatk1.0-0, libatk-bridge2.0-0, libgtk-3-0
+- libnss3, libcups2, libxss1, libgdk-pixbuf2.0-0, libasound2
+- nodejs, npm
+
+## User Detection
+
+The script automatically detects the appropriate user to run LM Studio:
+
+- If `LM_STUDIO_USER` is set and the user exists, it uses that user
+- If running as root, it checks for `ec2-user`, `ubuntu`, or uses `$SUDO_USER`
+- If running as a regular user, it uses the current user (`whoami`)
+- The detected user's home directory is used for LM Studio configuration
+
+## Troubleshooting
+
+### Common Issues
+
+1. **Service fails to start**: Check logs with `sudo journalctl -u lm-studio -f`
+2. **Permission denied**: Ensure the script is run with appropriate permissions (may require sudo)
+3. **CLI bootstrap fails**: The AppImage may need to be run interactively to install the CLI
+4. **Xvfb fails to start**: Check if Xvfb is installed and accessible
+5. **Download fails**: The AppImage is large (~1GB), ensure you have sufficient disk space and network bandwidth
+6. **Port already in use**: Check if another service is using the configured port (default: 1234)
+
+### Logs and Debugging
+
+- **Service logs**: `sudo journalctl -u lm-studio -f`
+- **Check if running**: `sudo systemctl is-active lm-studio`
+- **Check port**: `netstat -tlnp | grep :1234` or `ss -tlnp | grep :1234`
+- **API endpoint**: Access <http://localhost:1234/v1> (or your configured port)
+- **Bootstrap logs**: Check `/tmp/lmstudio-bootstrap.log` during installation
+
+### Manual CLI Bootstrap
+
+If the automatic CLI bootstrap fails, you can manually bootstrap the CLI:
+
+```bash
+# Start Xvfb
+sudo -u $USER Xvfb :99 -screen 0 1024x768x16 -ac -nolisten tcp &
+
+# Run AppImage briefly to install CLI
+sudo -u $USER DISPLAY=:99 /usr/local/bin/lm-studio --no-sandbox
+
+# Stop after a few seconds once CLI is installed (Ctrl+C)
+# CLI should now be available at ~/.lmstudio/bin/lms
+```
+
+### Download Issues
+
+The AppImage download is large (~1GB) and may take several minutes:
+
+- The script supports resume if download is interrupted
+- Partial downloads are preserved and resumed on retry
+- Check disk space: `df -h`
+- Check network connectivity: `ping installers.lmstudio.ai`
+
+### API Server
+
+The API server starts automatically when the service starts. To interact with it:
+
+```bash
+# Check if API is responding
+curl http://localhost:1234/v1/models
+
+# Start a chat completion (example)
+curl -X POST http://localhost:1234/v1/chat/completions \
+  -H "Content-Type: application/json" \
+  -d '{
+    "model": "your-model",
+    "messages": [{"role": "user", "content": "Hello!"}]
+  }'
+```
+
+**Note**: You need to download a model first using the CLI:
+
+```bash
+lms get <model-name>
+```
+
+## Security Considerations
+
+- **User permissions**: LM Studio runs as a non-root user
+- **Firewall**: Ensure the configured port is open in your firewall if accessing remotely
+- **Network security**: For production use, consider setting up a reverse proxy with HTTPS
+- **API access**: The API server is accessible on all interfaces (0.0.0.0) by default
+
+## Limitations
+
+- **Linux only**: LM Studio AppImage is only available for Linux
+- **Large download**: The AppImage is approximately 1GB in size
+- **GUI dependencies**: Requires GUI libraries even for headless operation (for AppImage bootstrap)
+- **Model downloads**: Models must be downloaded separately using the CLI
+- **Resource intensive**: Running LLMs requires significant CPU/GPU resources and memory
+
+## Additional Resources
+
+- **LM Studio Documentation**: <https://lmstudio.ai/docs>
+- **LM Studio GitHub**: <https://github.com/lmstudio-ai>
+- **API Documentation**: <https://lmstudio.ai/docs/api>