Skip to main content

Sessions

A session is your working environment: one container with a GPU slice (or a CPU-only allocation), an image, and optionally your volumes mounted into it. You open it in the browser with VS Code, JupyterLab, or a terminal — there is no SSH and nothing to install locally.

This section walks through the whole life of a session:

PageWhat it covers
Creating a sessionThe five-step wizard: compute, GPU, image, volumes, review
ConnectingVS Code, JupyterLab, the web terminal, and the single-use links
Managing a sessionPause, resume, restart, live usage, the event log
Waiting in the queueWhy a session waits, where it stands, how to cancel
Ending a sessionTerminating one or many, what is billed, what survives

The session list​

Session list

Only your own sessions are listed. Each row carries the state, the resources it holds, the sharing mode, GPU occupancy, uptime, and what it has cost so far.

List tabs and filters

The tabs split the list by state — Active (running, paused, waiting), Running, Terminated, and All — and the toolbar narrows it further by name or id, resource class, and GPU model. All of it lives in the address bar, so a filtered view survives a reload and can be pasted to someone else.

States​

StateWhat it meansBilling
PendingWaiting for capacity, or the container is startingCredits are held, nothing is consumed
RunningThe container is up and reachableBilled per second at rate × occupancy
PausedThe container is gone, the GPU is returned, the session and its volumes are keptNot billed
TerminatedFinished; the bill is settled and the unused hold refundedNothing further
ErrorRefused or failed to start — the reason is on the detail pageAny hold is released

Only GPU sessions consume credits. CPU sessions are free and are bounded by your resource policy instead.

The detail page​

Session detail

Everything about one session in a single screen: the resources it holds, where it runs, the image and volumes behind it, live usage, and the event log. The controls in the top-right are covered in Managing a session.