Deploy on Vast.ai
Vast.ai is a GPU marketplace where individuals and data centers rent out spare GPU capacity. It’s often the cheapest way to get a high-VRAM GPU.
1. Find an Instance
Log in to vast.ai and go to Search. Filter for your preferred GPU.
Look for instances with a Reliability score of 95%+ and a fast internet connection (500 Mbps+) for good model download speeds.
2. Configure the Instance
Click Rent on your chosen machine. In the configuration panel:
Image:
openlaboratoryorg/laboratory-osEnvironment Variables — set the following:
| Key | Value |
|---|---|
UPLINK_API_KEY | Your account key from openlaboratory.com. The lab connects through it and serves on your *.laboratory.computer URL. |
Disk: Allocate enough disk space for models. 50–100 GB is a reasonable starting point for a few image generation checkpoints. 200+ GB if you plan to download LLMs.
No inbound ports or firewall rules are needed — Laboratory OS uses an outbound tunnel. You can block all inbound traffic entirely.
3. Launch and Connect
Click Rent. The instance spins up and pulls the container image — this takes 1–3 minutes on first run depending on instance internet speed.
Check the container logs for your assigned slug — it appears within the first 30 seconds. Then sign in to your account at:
https://app.laboratory.computerFrom there you'll see your running instance and can open the Laboratory OS desktop. Access is gated by your SSO login rather than a shared token, so only authenticated account holders can connect.
4. Persistent Storage
Vast.ai instances reset their disk when destroyed. To persist data across instances:
- Use a Vast.ai volume (if available) mounted at
/data - Or export important models to external storage (S3, Google Drive, etc.) before destroying an instance
Note: Vast.ai instances can occasionally go offline if the host machine has issues. Save your work frequently and keep important models backed up.