Install Hermes on Linux

What this page covers

This page covers the standard Linux install path, including the official installer, shell reload, provider selection, and verification commands. It also applies to most WSL2 terminal setups after the Linux shell is working.

Hermes changes quickly, so treat the official documentation as the source of truth and use the built-in CLI to inspect the local install before editing config by hand.

Fast path

curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
source ~/.bashrc   # or source ~/.zshrc
hermes setup
hermes model
hermes chat -q "Say hello, then tell me which model you are using."
hermes doctor

Practical checklist

  • Install in the user shell that will actually run Hermes and the gateway.
  • Choose a provider with at least 64K context; the official docs note that Hermes rejects smaller-context models.
  • Use hermes setup --portal when a Nous Portal login should configure a model plus Tool Gateway tools in one flow.
  • If this will become a server/gateway machine, verify the normal CLI before installing background services.
  • For WSL2 gateway service work, ensure systemd support is configured before assuming persistent services will survive.

Common pitfalls

  • Debugging gateway or cron before the base CLI works.
  • Configuring secrets in config.yaml; secrets belong in .env or OAuth credential storage.
  • Using local/self-hosted models without confirming the context window.

Verification checklist

  • hermes chat -q ... returns a real response.
  • hermes doctor passes or reports only understood non-blockers.
  • hermes config shows the intended provider/model.
  • The target shell can run hermes without manually exporting PATH each time.

Official reference

https://hermes-agent.nousresearch.com/docs/getting-started/quickstart