Skip to main content
A job is any task that runs in the background, such as training, an evaluation, synthesis, an upload, an export, or a deployment. The Jobs page lists every job in the project, so you can see what is running, how far it got, and why a job failed. The API and CLI call a job an operation, and a job’s ID is its operation ID. FAiE (Frontier AI Engineer Agent) waits for the jobs it starts and reports the result in the conversation, so you don’t need to watch the page while FAiE works. Jobs page with the Type and Status filters and a list of jobs with their type, status, progress, and time elapsed

FOLLOW A JOB

The top bar shows how many jobs are running. Click it to see them, or open the Jobs page for the full list. Filter the list by Type or Status, or search by name. Click a job to open its details, including its ID, timestamps, progress, and result. A training job also shows its training progress.

WHEN A JOB FAILS

A failed job shows its error in the details. The error usually names the cause, such as a dataset row that doesn’t match the expected format, a missing permission, or a quota limit. Fix the cause and run the job again. A failed job doesn’t keep partial results. To get help with a failure, ask FAiE why the job failed, or include the job’s ID when you contact support.

STOP A JOB

Open a pending or running job, and click Stop. A stopped job doesn’t resume, so start a new job to run it again. Its recipe keeps the configuration for the next run.