Hermes Agent

Hermes Agent is an open-source command-line assistant from Nous Research. Connect it to this site as a custom endpoint.

• Protocol: OpenAI-compatible (Chat Completions)
• Base URL: https://tokenfunc.com/v1
• Models: models with 64K+ context, for example claude-sonnet-4-5

1. Install Hermes

   macOS / Linux

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

   Windows

   iex (irm https://hermes-agent.nousresearch.com/install.ps1)

2. Run the setup wizard

   hermes model

   Choose Custom endpoint and answer the prompts:

   Prompt	Value
   Base URL	https://tokenfunc.com/v1
   API key	sk-xxxxxxxxxxxxxxxx
   Model	claude-sonnet-4-5
   API mode	chat_completions
   Context length	The model's context length, for example 200000

   Hermes requires at least 64K of context. This site does not report context length, so enter it yourself.

   Screenshot pendingThe Custom endpoint screen of the hermes model wizardtools/hermes-wizard.png

3. Verify

   hermes chat -q "reply ok" --oneshot

   A reply means you are connected. hermes doctor also checks the configuration.

Configuring with a file

Instead of the wizard you can edit the config file and keep the key in a .env file next to it:

model:
  default: claude-sonnet-4-5
  provider: custom
  base_url: https://tokenfunc.com/v1
  key_env: TOKENFUNC_API_KEY

TOKENFUNC_API_KEY=sk-xxxxxxxxxxxxxxxx

model:
  default: claude-sonnet-4-5
  provider: custom
  base_url: https://tokenfunc.com/v1
  key_env: TOKENFUNC_API_KEY

TOKENFUNC_API_KEY=sk-xxxxxxxxxxxxxxxx

Switching models

In a session type /model custom:<model name>, for example /model custom:claude-sonnet-4-5. To add another endpoint, leave the session and run hermes model again.

Troubleshooting