A second account should mean more work, not more switching
Teams rarely have just one source of AI capacity. A founder may have a personal Claude plan and a company plan. A product team may use Codex for one repository and Gemini for another. Until now, reaching a limit often meant stopping the task, changing credentials, and starting again by hand.
That turns capacity into fragmentation. You technically have more of it, but you cannot use it as one dependable system. Grain now treats every connected account as available capacity while keeping the boundaries of each plan visible.
The same plan gets the next job
When an account reaches its limit, Grain first looks for another connected account on the same plan. Work moves there without changing the model or the way the task runs. Only after the matching accounts are unavailable does Grain consider another provider you have approved.
The result is simple: adding a second subscription increases throughput. It no longer adds another credential you need to remember to switch.
State you can inspect
Automatic routing only feels trustworthy when you can see what it is doing. The plan view now shows which accounts are active, which are limited, when capacity resets, and which need you to sign in again. Hosted Grain credits and the model assigned to each agent appear in the same place.
That makes the router legible. You can understand why a task moved, know what capacity remains, and make a deliberate change when the work calls for it.
What changes in practice
Overnight Missions can keep running after one account hits a cap. Personal and work subscriptions can contribute capacity without being collapsed into one identity. A temporary provider limit becomes a routing event instead of a failed run.
Grain still respects the machines, repositories, providers, and approval rules you choose. The goal is not to hide the system. It is to make the system do the coordination work for you.
- Connect more than one account for the plans your team already uses.
- See live availability and reset times from one view.
- Keep a Mission moving without manually restarting it on another provider.
- Preserve your preferred model for as long as matching capacity is available.
One workhorse, built from what you already have
The best infrastructure is often the infrastructure a team can adopt without replacing everything. Grain brings your subscriptions, local tools, repositories, and hosted models together behind one Mission. This release makes that promise more literal: every account you connect now makes the whole system stronger.
