DevelopersIntegrationsHermes Agent

Hermes Agent with ARK

Point Hermes at ARK as a custom endpoint and run it on open models with tool calling and a 256K-token context, well above the 64,000 tokens Hermes needs.

By Nous ResearchMIT licenceSourceVendor docs

Hermes Agent: The self-improving AI agent built by Nous Research.

Before you start

You need an ARK API key: sign in to ARK Cloud and copy it from the dashboard, or use a key from your own ARK deployment and its base URL. The setup below gives each of Hermes' model roles an ARK Cloud model: arklabsai/Qwen3.5-122B-A10B as the main model, arklabsai/Qwen3.6-35B-A3B for delegated subagents and arklabsai/Qwen3.5-4B for session titles. All three call tools; the 122B and the 35B read up to 256K tokens, the 4B up to 128K. Any chat model from the catalogue works.

1. Install

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

2. Connect it to ARK

Run hermes model and choose Custom endpoint, or set the models in ~/.hermes/config.yaml:

~/.hermes/config.yaml
model:
  default: arklabsai/Qwen3.5-122B-A10B
  provider: custom
  base_url: https://api.ark-labs.cloud/api/v1
  key_env: ARK_API_KEY
  context_length: 262144

delegation:
  model: arklabsai/Qwen3.6-35B-A3B

auxiliary:
  title_generation:
    model: arklabsai/Qwen3.5-4B
    base_url: https://api.ark-labs.cloud/api/v1
    api_key: ${ARK_API_KEY}

Put ARK_API_KEY=… in ~/.hermes/.env, where Hermes keeps secrets. Subagents run on delegation.model and, with no provider or base URL of their own, use the main endpoint and key. Hermes runs auxiliary tasks on the main model unless you give a task its own; auxiliary.title_generation sends session titles to the 4B.

Good to know

  • Use provider: custom with base_url; the OPENAI_BASE_URL variable applies only to Hermes' openai-api provider.
  • Hermes needs at least 64,000 tokens of context for agent use with tools.
  • Keep compression on the main model: the 4B reads up to 128K tokens, half the main model's window, and compression reads the whole conversation.
  • These tools send the whole conversation on every turn. ARK's prompt cache speeds up the repeated part of each prompt; every prompt token is still billed.

Next

Read about tool calling and limits and capacity on ARK, or browse other integrations.

Configuration keys follow Hermes Agent's official documentation as of November 2026. If the tool has changed since, its docs take precedence.