selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

hermes-agent logo

Self-host hermes-agent

v0.23.1

AI agent that learns, creates skills, remembers

BetaOpen SourceMIT
Updated6 days ago
LanguagePython
Setup difficultyModerate
ThemeDark mode
Stars
263 493
Rating
5.0 (1)
Ports
8642 (OpenAI-compatible API server), 8443 (Telegram webhook), 8090 (WhatsApp Cloud webhook), 8644 (Generic webhook platform), 8646 (Microsoft Graph / LINE webhook), 3978 (Microsoft Teams (Bot Framework) webhook), 8645 (WeCom / BlueBubbles webhook)
Docker
nousresearch/hermes-sandbox:desktop
Install with Linux
Messaging platform connections overviewPersistent memory featureNatural language cron schedulerIsolated subagents for parallel workWeb search and browser automationIsolated sandboxing with multiple backendsHero screenshot of Hermes Agent

What is hermes-agent?

Hermes Agent is a self-hosted AI assistant from Nous Research. It builds a persistent memory that grows across conversations, automatically generates reusable skills from completed tasks, and refines those skills over time. The software runs as a terminal interface with multiline editing and conversation history, plus native desktop applications for macOS, Windows, and Linux. A single gateway process connects the agent to multiple messaging platforms while sharing one memory store. Users can choose any LLM provider — OpenRouter, Nous Portal, OpenAI, Anthropic, Google, local models via Ollama or vLLM, or custom OpenAI-compatible endpoints — and switch between them at runtime. Features include a natural-language cron scheduler, isolated subagents for parallel work, seven terminal execution backends, web search, browser automation, vision analysis, image generation, and text-to-speech. All conversation data, memories, and skills remain on the user's machine unless the optional Nous Portal cloud subscription is enabled.

Should you self-host hermes-agent?

hermes-agent scores 89/100 on selfhost.directory (how scoring works) and ranks #1 of 120 in AI Assistant. It has 263 493 GitHub stars, +10 670 in the last 30 days.

It shipped 40 stable releases in the past year. The latest, v0.23.1, came out on 10 November 2026. New versions land roughly every 4 days.

Setup is moderate: expect some configuration beyond a single command. An official Docker image is available. It is MIT-licensed and written mostly in Python.

Its closest alternative on selfhost.directory is AnythingLLM (87/100, 68 294 stars). hermes-agent scores higher overall. See hermes-agent vs AnythingLLM side by side.

Features

  • Autonomous Skill Creation
  • Context Compression
  • Isolated Subagents
  • MCP (Model Context Protocol) Support
  • Multi-Platform Messaging Gateway
  • Multiple Terminal Backends
  • Natural Language Cron Scheduler
  • Persistent memory
  • Session Search
  • Voice Memo Transcription

hermes-agent README

How to install hermes-agent

Install hermes-agent with Linux

Environment variables
HERMES_HOME, TERMINAL_ENV, etc.

Native installation via installer script

hermes-agent releases & version history

@ Every self-hosted release, weekly

One Monday email with the week's stable releases across the directory.

FAQ

How much does it cost?

Hermes Agent itself is free and open-source (MIT license). You pay only for LLM API usage from your chosen provider. Local models are completely free to run.

Is my data sent anywhere?

API calls go only to the LLM provider you configure. Hermes Agent does not collect telemetry, usage data, or analytics. Conversations, memory, and skills are stored locally in ~/.hermes/.

More in AI Assistant

Rate this project

Click to rate — no account needed.

0 comments

Posting as guest
Loading comments…