Remote GPU
Move inference, not your workspace.
Connect TuxNest to an Ollama server on a GPU machine you own or control. The extension tunnels the model API over SSH while local file edits and normal commands stay on your development machine.
Connection flow
Verify the host before you trust it.
Prepare the host
Install and run Ollama on the remote machine yourself. TuxNest does not automatically start arbitrary remote services.
Configure SSH
Run TuxNest: Configure Remote GPU Host and provide your host, user, authentication method and remote Ollama port.
Confirm identity
Verify the remote host fingerprint independently before accepting it. Treat host-key trust as a security decision, not a cosmetic prompt.
Connect and discover
Run TuxNest: Connect to Remote GPU Host. TuxNest tunnels the API and registers models actually discovered on that host.
Observe lifecycle
If the SSH or studio connection closes, remote models are removed. TuxNest does not silently replace the selected model with another provider.
Important boundary
The GPU is for inference.
Connecting a remote model does not turn the GPU host into your workspace machine. Your approved local file operations and commands remain scoped to the local development environment unless you explicitly use a different mechanism.
TuxNest hybrid compute
Your GPU. Your AI. Your Control. Anywhere.
TuxNest connects a reviewable VS Code coding workflow with selected local models, SSH-tunneled models on user-controlled GPU workstations and explicitly configured compatible endpoints.