Hermes

Connect Hermes to the DeepSeek API

Author
DeepSeekAgent.io Editorial Team
Published
Updated

Hermes is a self-improving AI agent built by Nous Research. It includes a built-in learning loop: it creates skills from experience, improves them during use, persists knowledge, and builds an evolving model of your preferences across sessions.

1. Install Hermes

Quick Install

Get Hermes Agent up and running in under two minutes with the one-line installer.

Linux / macOS / WSL2
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash

The only prerequisite is Git. The installer automatically handles everything else.

For more installation instructions, please refer to the Hermes installation page.

2. Configure a DeepSeek model

After installation, reload your shell and run the full model setup wizard from a regular terminal:

hermes model

hermes model is the current entry point for adding providers, entering credentials, and selecting a model. Choose DeepSeek and enter your DeepSeek API Key. Hermes writes the provider configuration for you, so there is no need to preserve an old hard-coded model ID.

You can also put the key in the Hermes environment file before opening the wizard:

mkdir -p ~/.hermes
printf 'DEEPSEEK_API_KEY=%s\n' 'your API key' >> ~/.hermes/.env
hermes model

Do not commit a real key to a project repository. Once configuration is complete, start Hermes with:

hermes

The in-session /model command only switches among providers and models that are already configured. To add DeepSeek for the first time or change its endpoint and credentials, exit the session and run hermes model.

3. Use a compatible API endpoint

For a DeepSeek-compatible API rather than Hermes' built-in DeepSeek provider, choose Custom endpoint in hermes model, then enter the service's Base URL, API key, and model ID. Use the model name exposed by that service instead of copying the old deepseek-v4-pro value.

If you need a compatible API key for testing, visit the free DeepSeek API credits page to claim a key and see the currently supported models.

4. Quick checks

  • To add or reconfigure a provider, run hermes model outside a session.
  • To switch among configured models, run /model inside a session.
  • If an environment key is not detected, confirm it is stored in ~/.hermes/.env, then restart Hermes.
  • If a custom service cannot connect, verify the version path in the Base URL and the exact model ID exposed by the service.

References