Portainer Templates logo

Portainer Templates

Lm-studio Lm-studio

Container

Remote DesktopAI

LM Studio can run local AI models like gpt-oss, Llama, Gemma, Qwen, and DeepSeek privately on your computer.

Image details

Pulls: 10.7k
Architecture: amd64
Image size: 8.2 GB
User: linuxserver
Created: Jun 26, 2026
Updated: 1 day ago
Status: active

Configuration

Type
Container
Platform
linux
Image
linuxserver/lm-studio:latest
Ports
3000:3000/tcp3001:3001/tcp
Volumes
/config : /srv/lsio/lm-studio/config
Env vars
PUID=1000PGID=1000TZ=Etc/UTC
Restart
unless-stopped

Template by technorabilia

Notes

Portainer App Templates by Technorabilia, based on data provided by LinuxServer.io.

Ensure to create the following volume directories on the host file system, or modify the paths in the volume mapping section under the advanced options below, as needed.

mkdir -p /srv/lsio/lm-studio/config

Standalone Install

Select an install method, to see config/commands for deploying Lm-studio

Installation method

Install on Portainer

Import all app templates into your Portainer instance, for easy 1-click deploys

  1. Ensure both Docker and Portainer are installed, and up-to-date
  2. Log into your Portainer web UI
  3. Under Settings → App Templates, paste the below URL
  4. Head to Home → App Templates, and the list of apps will show up
  5. Select Lm-studio, fill in any config options, and hit Deploy

Template Import URL

https://raw.githubusercontent.com/Lissy93/portainer-templates/main/templates.json
Show Me demo

More install options in our documentation.

linuxserver/lm-studio

This readme has been truncated from the full version found HERE

LM Studio can run local AI models like gpt-oss, Llama, Gemma, Qwen, and DeepSeek privately on your computer.

Application Setup

LM Studio can be accessed at:
  • https://yourhost:3001/

GPU Support

This image is Arch based and supports the following GPUs:
  • Cuda v13- This requires a 2000 series or higher Nvidia video card running the latest binary drivers, v12 is not supported.
  • Vulkan- This requires a semi modern AMD GPU.

ROCm is not supported at this time due to it's size and Vulkan option for AMD.

Minimum run commands (start here then expand to fit your needs)

AMD GPUs:
docker run --rm -it \
  --shm-size=1gb \
  --device /dev/dri \
  -e AUTO_GPU=true \
  -p 3001:3001 \
  linuxserver/lm-studio bash

Nvidia GPUs:
As of driver 595.80 with NVIDIA Container Toolkit installed:
# Host level modprobe modeset
nvidia-modprobe --modeset

docker run --rm -it \
  --shm-size=1gb \
  --device /dev/nvidia-modeset \
  --runtime nvidia \
  --gpus all \
  -e AUTO_GPU=true \
  -p 3001:3001 \
  linuxserver/lm-studio bash

AI tools included in this image

This image comes preloaded with a set of AI development and automation tools centered around LM Studio, enabling local model execution, and integration inside a KDE desktop environment.
At the core is LM Studio along with its developer ecosystem via lmstudio SDK, providing programmatic access to local models and workflows.
For AI-assisted coding and agentic workflows, the image includes:
  • Aider (via aider-install) for git-aware code editing with LLM assistance (manual configuration required)
  • Cline for agent style development inside development environments
  • Code-OSS for a desktop IDE with AI plugin support
  • OpenClaw for running autonomous AI agents that execute tasks through workflows and messaging based interaction
  • OpenCode AI for code generation and refactoring

These tools are primarily CLI and developer focused, and are designed to be configured by the user from the terminal on first launch to connect to a local LM Studio backend.
Included is a desktop launcher for LLMster to provide the backend for these tools in the desktop environment in a headless manner. This can also be launched with lms server start from the command line or from the desktop application via the taskbar icon.
Important notes:
  • LM Studio (the desktop application) and LLMster cannot be run at the same time, as they both manage overlapping runtime resources for local model access and will conflict if launched concurrently.
  • All included AI tools require initial configuration to point to the active LM Studio backend before use.

Usage

docker run -d \
  --name=lm-studio \
  -e PUID=1000 \
  -e PGID=1000 \
  -e TZ=Etc/UTC \
  -p 3000:3000 \
  -p 3001:3001 \
  -v /path/to/config:/config \
  --shm-size="1gb" `#optional` \
  --restart unless-stopped \
  lscr.io/linuxserver/lm-studio:latest

Serve Lm-studio on your own domain behind Caddy, Nginx or Traefik. Fill in your domain and copy the result. It's a starting point, some apps need their own base URL or extra headers set too.

Proxying lm-studio.example.com to http://Lm-studio:3000

Add this to your Caddyfile

lm-studio.example.com {
	reverse_proxy http://Lm-studio:3000
}

Check the logs first

Nine times out of ten the logs tell you exactly what went wrong.

  • In Portainer, go to Containers, click the container, then Logs. Or run docker logs Lm-studio
  • Exit codes help too: 137 means killed, usually out of memory. 126 or 127 means the command inside the image is broken.

Port already in use

If deployment fails with "Bind for 0.0.0.0:3000 failed: port is already allocated", something else on your server is using that port.

  • Find what's using it: sudo ss -tlnp | grep :3000
  • Stop the other service, or pick a different host port. In 3000:3000 only the left number is yours to change, the right one belongs to the app.

Running but the page won't load

The container is up but nothing appears in your browser.

  • Use your server's real IP: http://your-server-ip:3000. The 0.0.0.0 link Portainer shows isn't a real address.
  • Give it a minute after first deploy, Lm-studio can take a while to initialise.
  • Make sure your firewall allows the port, e.g. sudo ufw allow 3000

Permission denied on volumes

If the logs show "permission denied", the app can't write to its data folder on the host.

  • Fix the ownership: sudo chown -R 1000:1000 /srv/lsio/lm-studio/config
  • Or set the PUID and PGID variables (defaults 1000:1000) to match your own user, found with id $USER

Image won't pull

Test the pull directly on the host: docker pull linuxserver/lm-studio:latest

  • "manifest unknown" means the tag no longer exists. This template uses latest, so try pinning a specific version instead.
  • "toomanyrequests" is the Docker Hub rate limit. Log in with docker login to raise it.
  • "no space left on device" means a full disk. Reclaim space with docker system prune

"exec format error"

This means the image was built for a different CPU architecture than your server.

  • This image supports: amd64
  • Check yours with uname -m: x86_64 is amd64, aarch64 is arm64. Raspberry Pi and other ARM boards are the usual culprits.

Container keeps restarting

The unless-stopped restart policy relaunches the app after every crash, so the real error can scroll past.

  • Check the logs right after a restart, the last few lines before it died are the useful ones.
  • Get the exit code with docker inspect Lm-studio --format '{{.State.ExitCode}}'
  • Still stuck? Redeploy once with the restart policy set to no so the failure stays visible.

Raise an issue

Found something which isn't working as it should? Here's how to report it.

A single container

Lm-studio runs as one container, the simplest kind of app here. Just the one image to pull and nothing else wired up alongside it.

The app image

An image is the app packed up ready to go, everything Lm-studio needs bundled into one download. This template pulls linuxserver/lm-studio:latest, which Docker fetches once (about 8.2 GB) and then starts your own copy from.

Where the image comes from

Docker pulls its images from registries, public libraries of ready-built apps. Lm-studio's comes from Docker Hub, published by linuxserver.

Version tags

The bit after the colon in the image name is the version tag. Here it's latest, which always points at the newest build, so a redeploy can bump you to a newer release without you asking. Pin a specific tag if you would rather stay on one version.

Which machines it runs on

Every image is built for particular CPU types. This one ships for amd64, so it runs on regular x86 PCs and servers, though not ARM boards like a Raspberry Pi.

Ports

A port is the door the app answers on. A mapping like 3000:3000 means it's reachable on port 3000 of your server, where the left number is yours to change and the right one belongs to the app. Once it's running, open http://your-server-ip:3000 in a browser. It opens:

  • 3000:3000, likely the web interface
  • 3001:3001

Volumes

A volume is where Lm-studio keeps its files so they survive an update or a restart. Without one, anything it saves would sit inside the container and vanish the moment it's recreated. This template mounts:

  • /config from /srv/lsio/lm-studio/config on the host

Environment variables

Environment variables are the settings you hand over when you deploy, things like a password or a timezone. Lm-studio takes 3 of them, all with defaults you can leave alone or tweak:

  • PUID, defaults to 1000. Run 'id [USER]' for the owner of the host volume directories to get the UID to use here.
  • PGID, defaults to 1000. Run 'id [USER]' for the owner of the host volume directories to get the GID to use here.
  • TZ, defaults to Etc/UTC. Specify a timezone to use, see this list.

Restart policy

The restart policy here is unless-stopped, so Docker restarts Lm-studio after a crash or reboot, but leaves it off when you stop it on purpose. You can change this on the deploy screen. The choices are no (never restart), on-failure (only after a crash), unless-stopped (restart unless you stop it), and always (bring it back no matter what).

Users and permissions

The PUID and PGID settings tell it which user and group to act as on your host. Point them at your own account (find yours with id $USER) so the files it writes into your mounted folders come out owned by you rather than root.

Networking

Nothing custom is set, so Lm-studio sits on Docker's default bridge network: its own private space that reaches the outside world only through the ports it publishes.

Container name

Once it's deployed, Portainer names the container Lm-studio. That's what you'll spot in the containers list and use in commands like docker logs Lm-studio.

Platform

The platform is linux, the kind of system the container is built to run on. Docker and Portainer handle this on a normal Linux server.

Portainer app templates

Zooming out, this whole page comes from a Portainer app template: a short recipe telling Portainer how to set Lm-studio up. Add the template list to Portainer once, then deploying Lm-studio is a click rather than a wall of config.