All Apps
comfyanonymous /
ComfyUI
Generate images and videos using a powerful low-level workflow graph builder - the fastest, most flexible, and most advanced visual generation UI.
Create powerful and fast image generation pipelines
Lab Station OS
The Linux Server for Open-Source AI
Workstation simplicity with the power of cloud GPUs. Use the web desktop to install apps, download models, run advanced workflows, and deploy self-hosted APIs.
Run
100+
models from the world's top AI research labs.
Install
10+
self-hosted AI research tools and apps.
Deploy
5
of the most powerful Nvidia enterprise GPUs.
Powerful Tools
For Researchers and Engineers
Lab Station OS is pre-configured for the latest Nvidia hardware and AI frameworks. Stop wasting time wrangling GPU drivers, CUDA libraries, and Python environments.
Easy Model Management
Universal Compatibility
Compatible with all AI model architectures and file formats.
Upload
Add models and data from your local computer via encrypted file transfer using the File Browser app.
Download
Download models from Hugging Face, Civitai, Ollama, or directly from any URL - at speeds up to 10 Gbps.
Browse
One-click download the most popular and capable open-source models, directly from the built-in Model Directory.
Fine Tune
Fine-tune models with your own datasets using efficient techniques such as LoRA and PEFT.
Secure, Private, Powerful
Your Server, Under Your Control
Dedicated Linux VMs with Nvidia GPUs.
Enterprise datacenter security and reliability.
Powerful GPU Servers
ready to deploy
One-click deploy private Lab Stations onto datacenter servers with the latest Nvidia AI accelerators and enterprise-grade security.
On-Demand
Pay for what you use, prorated to the second - no mandatory commitments or subscriptions.
Fast Networking
Ultra-fast datacenter fiber optic network speeds - download large models onto the Lab Station in seconds.
Enterprise GPU Servers
Utilize the latest cutting-edge Nvidia hardware, run your workloads on the most powerful AI accelerators.
Competitive Pricing
Savings of 40%+ compared to deploying the same GPUs on public clouds such as AWS, GCP, and Azure.
Nvidia
A5000
Lab Station GPU Server
Nvidia
A40
Lab Station GPU Server
Nvidia
L40S
Lab Station GPU Server
Nvidia
A100
Lab Station GPU Server
Nvidia
H100
Lab Station GPU Server
Persistent Lab Stations
Save Time, Reduce Costs
Lab Stations can be paused whenever you need to step away. The full system state is saved so that you can pick up where you left off.
Unlimitted Pause/Resume
Setup once, use whenever - the convenience of a local workstation with the power and flexibility of the cloud.
Persistent System State
Not just a mounted volume, the full boot disk is saved.
Equivalent to turning off a local workstation computer.
Fast NVMe Storage
Physically connected storage with low I/O latency.
Efficiently manage massive datasets and model weights.
Cost Savings
Reduce costs by pausing your server when not in use.
You do not pay for the GPU when the system is paused.
Storage Cost
Persistent Lab Stations
Save Time, Reduce Costs
Lab Stations can be paused whenever you need to step away - the full system state is saved, including the boot disk, so that you can pick up where you left off.
Open Source Moves Fast
Stay on the Cutting Edge
Explore the latest advances in AI research.
Run the models and apps on a dedicated cloud GPU server.