Skip to main content

Hermes Agent

Run Hermes Agent on your Standard Compute budget. Hermes runs on macOS, Linux and WSL. On Windows, run every command below inside WSL.

1. Install Hermes and the launcher​

Install Hermes by following its installation guide. Then install the launcher:

curl -fsSL https://standardcompute.com/cli/install.sh | sh

The launcher requires Node.js 20 or newer. Open a new terminal after installing.

2. Launch Hermes​

standardcompute launch hermes

Choose an API key and approve the connection in your browser. Hermes opens automatically.

How the connection works​

The launcher uses a separate, saved Standard Compute profile with its own sessions. Your usual Hermes setup and any connected messaging services stay untouched. Run hermes directly to return to them.

The profile points the main model and all auxiliary model slots (vision, compression, title generation and so on) at standardcompute over Chat Completions.

note

External tools such as web search or image generation may need their own credentials. The launcher opens the terminal agent, not a messaging gateway.

Manual configuration​

To configure Hermes yourself, run hermes setup, choose Full setup, and select Custom OpenAI-compatible endpoint as the provider:

Base URL:  https://api.stdcmpt.com/v1
API key:   YOUR_API_KEY
Model:     standardcompute

Troubleshooting​

  • Looping or picking the wrong tool: enable fewer tools, write clearer instructions, or lower max_turns. hermes doctor diagnoses most setup problems.
  • Rotated your key or switching accounts: run standardcompute login.
  • Unexpected request errors: re-run the installer to update the launcher.