How it fits together
Four things, and it helps to keep them apart.
Credit pays for the chat's model requests; the plan sets the container's hardware and only tops up credit.
Diagram as text
- A chat is a conversation with one model at a time, and you can change the model.
- A chat can be attached to a container. One container can serve several chats, and its files persist.
- Each answer is a model request that is paid per request.
- Credit, prepaid money, pays for those requests.
- The plan sets the container's limits: cores, memory, disk, how many containers run, how long an agent may run and how many public ports.
- Every paid plan also adds some credit each month. The plan itself does not pay for model requests.
The four things
Models
Hardware
You use
Models
Chat
The conversation. It belongs to you.
one model at a time, changeable
Hardware
Container
A real Linux machine, independent of any chat. You can attach the same
container to several chats.
its files persist
You pay
Models
Credit
Prepaid money for model usage. Separate from the plan, though every paid plan adds some each month (see Plans and Credit).
paid per request
Hardware
Plan
What hardware you may use.
- cores
- memory and disk
- how many containers run at once
- how long an agent may run
- how many public ports
The important consequence
Plan
Model usage
A plan does not pay for model usage. It only adds some credit each month.
Credit
Hardware
Credit does not buy hardware.
They are two meters because they are two different costs.
What survives closing the tab
Keeps running
- A chat, an agent or a /goal run once you have sent it
- A terminal session you can reattach to later
- A container's files, across restarts
Needs your action
- Stopping a run: the stop button or an explicit abort
- Nothing stops it just because the tab closed or the page reloaded
See Isolation for what else does and does not
change about a workspace.