Can I run multiple AI agents at the same time in Draftbit?

Yes, within your plan’s limit. Free and Standard run one agent at a time. Pro runs up to 5 at once, and Enterprise has no fixed limit. Each agent works in its own chat thread, so running several means opening several threads, and each thread can use a different agent and model. Inside a single thread only one run happens at a time. If you send another prompt while the agent is busy, Draftbit queues it and runs it after the current run finishes. The Tasks page is stricter and waits for a running thread in the same app to finish before it starts another task.

A related limit is how many projects can be open at the same time. A project is open while any of its apps has a running sandbox, and every app in an open project can run. Free allows one open project, Standard one plus one per added seat, Pro 3 plus one per added seat, and Enterprise is unlimited. Threads inside one app share that app’s sandbox, so this limit only matters when you want agents working in apps from several projects at once.

Parallel work is safest when each thread owns a distinct part of the project. Two agents editing the same navigation file or shared component at the same moment can overwrite each other, because they work on the same files in the same sandbox. Give each thread a narrow goal and name the files or flows it should leave alone. A Read Only thread is a safe way to have one agent investigate while another edits.

More agents do not make one feature finish sooner when its pieces depend on each other. A feature with a shared data model and several dependent screens is usually easier to build in sequence, reviewing one result before starting the next.