HF Jobs supports parallel runs, cron schedules, Hub webhooks, retries, logs and resource metrics. Webhook launches have a current limitation: they do not keep the original Job's volume mounts.

Jobs can be launched individually or in parallel. Scheduled Jobs accept cron expressions and aliases such as @hourly and @daily.

Hub webhooks can start work when watched Hugging Face repositories or buckets change. This is a native Hub integration; another system can submit a Job through the HTTP API or CLI.

The Python client supports automatic retry attempts and rerunning a saved specification. CLI commands and the Hub show status and logs, while hf jobs stats reports CPU, RAM, network and GPU usage.

A relevant limitation in the current webhook documentation: webhook runs do not retain the original Job's mounted volumes. For a UV script, the documented workaround is a Docker Job that runs uv run <url>, rather than relying on the volume created by hf jobs uv run.

Sources: Scheduling, webhooks, management, retries.