Hermes Agent

Hermes Agent supports custom OpenAI-compatible endpoints.

Set it up

Run hermes setup, choose Custom OpenAI-compatible endpoint, and enter:

Setting	Value
Base URL	`https://api.inferray.com/v1`
API key	Your Inferray key
Model	A model ID from the catalogue, such as `deepseek-v4-flash`

Or write the files yourself. In ~/.hermes/config.yaml:

model:
  provider: custom
  default: deepseek-v4-flash
  base_url: https://api.inferray.com/v1

And the credentials in ~/.hermes/.env:

OPENAI_API_KEY=inferray_...
OPENAI_BASE_URL=https://api.inferray.com/v1

Keep the .env file private.

cd your-project
hermes chat

For a single task that exits when done:

hermes chat --oneshot -q "Fix the failing test and run it again"

--model switches the model for a session.