Hermes One

// Overview

Hermes One is a community-maintained native desktop application for Hermes Agent, the self-improving assistant from Nous Research. Instead of driving the agent's CLI by hand, the app walks you through install, provider setup and day-to-day usage in one window. It uses the official Hermes install script, keeps the agent in `~/.hermes`, and puts a GUI in front of chat, sessions, profiles, memory, skills, tools, scheduling and messaging gateways.

On first launch it asks whether you want to run Hermes locally or connect to a remote Hermes API server. In local mode it checks `~/.hermes`, runs the official installer and resolves dependencies (Git, uv, Python 3.11+), then talks to the agent over `http://127.0.0.1:8642` with SSE streaming. In remote mode it validates a URL plus API key and skips the local install entirely, using the same streaming protocol.

The chat surface renders tool progress, markdown and syntax highlighting as the stream arrives, tracks prompt and completion tokens with a live cost readout, and exposes 22 slash commands (`/new`, `/web`, `/browse`, `/code`, `/shell`, `/compact`, `/model`, `/memory`, `/persona` and more). Sessions are searchable with SQLite FTS5, profiles isolate separate Hermes environments, and the Settings screen carries backup/restore, a log viewer and a debug dump. It is MIT-licensed, built on Electron with electron-updater for auto-updates, and ships an English locale with the i18n framework wired for community translations.

// What it does

• Guided first-run install of Hermes Agent with progress tracking and dependency resolution (Git, uv, Python 3.11+)
• Local backend on 127.0.0.1:8642 or a remote Hermes API server reached with URL + API key
• Multi-provider support — OpenRouter, Anthropic, OpenAI, Google, xAI, Nous Portal, Qwen, MiniMax, Hugging Face, Groq, plus local OpenAI-compatible endpoints (LM Studio, Ollama, vLLM, llama.cpp)
• Streaming chat with tool progress, markdown, syntax highlighting, live token counts and 22 slash commands
• Session management with SQLite FTS5 full-text search and date-grouped history; profile switching with isolated config
• 14 toolsets (web, browser, terminal, file, code execution, vision, image gen, TTS, skills, memory, session search, clarify, delegation, MoA, task planning)
• Memory browser with discoverable providers (Honcho, Hindsight, Mem0, RetainDB, Supermemory, ByteRover) and a SOUL.md persona editor
• Cron scheduler with 15 delivery targets and 16 messaging gateways including Telegram, Discord, Slack, WhatsApp, Signal, Matrix, Email, SMS and Home Assistant

// Getting started

Hermes One ships prebuilt installers for macOS, Windows and Linux from hermesone.org and the GitHub releases page. It installs Hermes Agent for you on first run, so no prior CLI setup is needed.

sudo dnf install ./hermes-desktop-<version>.rpm

/tools      # enable or disable toolsets
/skills     # browse and install skills
/memory     # inspect and edit memory entries
/usage      # prompt + completion tokens and cost
/compact    # compress the running conversation

Commands and code are distilled from the project's own documentation — always check the official repo for the latest.

// When to use it

• Reach for it when you want Hermes Agent running on a laptop without managing its CLI, installer and config files by hand
• Reach for it to give a self-hosted agent a chat UI with session search, persona editing and visible tool progress
• Reach for it when the agent should answer from messaging apps — Telegram, Discord, Slack, WhatsApp, Signal or email — through one configuration screen
• Reach for it to run scheduled agent jobs and have the results delivered somewhere you already read

// How Hermes One compares

Hermes One alongside other open-source assistants & chatbots tools AI/TLDR tracks, ranked by GitHub stars.