Hermes

Open-source autonomous AI agent with persistent memory. Connect it to ARZASTORE AI through a custom OpenAI-compatible endpoint.

Custom OpenAI-Compatible Endpoint

About Hermes

Hermes Agent is an open-source autonomous agent by Nous Research. It lives on your server, remembers what it learns and works with any inference provider - including a custom OpenAI-compatible endpoint like ARZASTORE AI.

Remembers context across sessions and grows more capable over time.

Full terminal UI with streaming output and slash-command autocomplete.

Connect to any provider - no lock-in, switch with one command.

Terminal, file operations, web search, memory and skills built in.

Installation

Install Hermes Agent with the official script:

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

Verify the installation:

hermes --version

Configuration

Hermes guides you through the model connection method. Choose the custom OpenAI-compatible endpoint option to use ARZASTORE AI:

Connect ARZASTORE AI

01Run hermes model to open the model connection wizard.
02When asked for the connection method, choose Custom OpenAI-compatible endpoint.
03Enter your ARZASTORE AI details when prompted:TEXTCopyBase URL: https://api.arzastore.com/v1 API Key: sk-your-api-key Model: claude-sonnet-4-20250514 Context window: 200000
04Confirm and start chatting - Hermes will route every turn through ARZASTORE AI.

Base URL: https://api.arzastore.com/v1
API Key: sk-your-api-key
Model: claude-sonnet-4-20250514
Context window: 200000

Switch anytime

Usage

Terminal UI

Start the terminal UI and chat with your agent backed by ARZASTORE AI:

hermes

Local Proxy

Expose a local OpenAI-compatible endpoint so other tools (Codex CLI, Aider, Cline...) can reuse the same backend:

hermes proxy