Skip to main content
A terminal AI agent from Nous Research (command hermes). Site: https://hermes-agent.nousresearch.com (repo: https://github.com/NousResearch/hermes-agent). Supported on Linux / macOS / WSL2. Windows is not natively supported — install inside WSL2.
Protocol: OpenAI-compatible (Chat Completions)
  • Base URL: https://api.orcarouter.ai/v1 (must include /v1)

​
Install

On Linux / macOS / WSL2: 
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
source ~/.bashrc   # or ~/.zshrc
On Windows, Hermes is not natively supported — open a WSL2 (Ubuntu) shell and run the same command above inside it.

​
Configure

Edit ~/.hermes/config.yaml: 
model:
  api_mode: chat_completions
  base_url: https://api.orcarouter.ai/v1
  default: orcarouter/auto
  provider: custom

​
Run

Hermes is interactive. Launch it and verify the model responds: 
hermes          # or: hermes --tui
Inside the session, use /model to switch models. You can also run hermes model first to configure a Custom Endpoint via the interactive menu.

​
Notes

  • The base URL keeps the trailing /v1.
  • Use api_mode: chat_completions — OrcaRouter speaks Chat Completions, not the Responses API.
  • default: orcarouter/auto uses automatic routing; provider: custom selects the custom endpoint.
  • Model names use vendor/model format. Swap orcarouter/auto for a specific model like anthropic/claude-opus-4.8 or google/gemini-2.5-flash.
  • On Windows you must run Hermes inside WSL2.