Start a managed Hermes workspace

Before you start

• Create an ApiClo account and review the currently offered Hermes package and hosting price.
• Check the task's model usage tariff separately from the workspace's recurring hosting charge.

Steps

1. Choose a workspace

   Open the managed agent section and select the available Hermes package that fits your workload.

2. Review and start provisioning

   Confirm the current infrastructure, monthly hosting charge and account balance before starting the guided deployment. Provisioning can take time and may require follow-up setup.

3. Run a small task

   Wait for the workspace to report readiness, then try a short task and inspect its status, files and model usage before a longer job.

Check the result

Confirm the workspace status is ready and one small task finishes with the expected output. Review hosting and token-usage records separately; starting deployment is not proof that every plugin or local integration works.

Troubleshooting

• Provisioning is pending or failed: inspect the workspace status rather than repeatedly creating new workspaces.
• An external account, plugin or local tool is missing: follow its separate connection and permission flow.
• A task uses paid models: check the model tariff and balance independently of the Hermes hosting price.

Compatibility and billing

Hermes is a managed agent workspace, not a free ChatGPT replacement. Hosting has a separate recurring charge, and model usage is billed according to current tariffs. Availability, package contents and supported integrations can change; ApiClo is independent of OpenAI and Anthropic.