# Hermes Agent — card for AI agents

Open-source agent from Nous Research, described by its makers as self-improving, that keeps persistent memory, creates and refines skills from experience, runs scheduled automations, and works from a terminal interface or messaging apps such as Telegram, Discord, Slack and WhatsApp. It can run locally or on Docker, SSH, Modal, Daytona and Vercel Sandbox backends, and has a desktop app.

Maker: Nous Research · category: General-purpose AI agents · open source (MIT)

## Price
- Free and open source (MIT); users pay their chosen model providers (Nous Portal, OpenRouter, OpenAI or custom endpoints).
- Free tier: yes
- Pricing page: https://github.com/NousResearch/hermes-agent
- Checked: 2026-11-05

## Install
- PyPI: `pip install hermes-agent`
- Homebrew: `brew install hermes-agent`
- Docker: `docker pull nousresearch/hermes-agent`

## Facts
- Interfaces: CLI, Messaging apps, Desktop
- Platforms: macOS, Linux, Windows, Android
- Bring your own API key: yes
- Repository: https://github.com/NousResearch/hermes-agent

## Numbers (AgentGid data)
- Gid Score: 63 (rank #22) (public popularity and activity, not quality)
- Downloads last 7 days: 88,207
- GitHub stars: 264,558

## Details (plan limits, live status, safety; from AgentGid's data)
Hermes Agent by Nous Research. Category: General-purpose AI agents. Type: personal assistant agent. Open-source agent from Nous Research, described by its makers as self-improving, that keeps persistent memory, creates and refines skills from experience, runs scheduled automations, and works from a terminal interface or messaging apps such as Telegram, Discord, Slack and WhatsApp. It can run locally or on Docker, SSH, Modal, Daytona and Vercel Sandbox backends, and has a desktop app. Open source (MIT) Pricing (checked 2026-11-05): Free and open source (MIT); users pay their chosen model providers (Nous Portal, OpenRouter, OpenAI or custom endpoints). Free tier: yes. Interfaces: CLI, Messaging apps, Desktop Platforms: macOS, Linux, Windows, Android Models: Any model: Nous Portal, OpenRouter, OpenAI and custom endpoints; a Claude Pro or Max plan through an experimental plugin Supports: mcp, multi_agent, headless Data: Gid Score 63/100, #22 overall; #3 in its category; 264,558.0 GitHub stars; latest release 0.23.2 (2026-11-16). News: 2026-06: Hermes Desktop app launched (per Wikipedia and press cited there). News: Claude Pro or Max plan instead of an API key (experimental): Nous Research's official plugin Claude Subscription DirectSDK (v0.3.4) runs Hermes through the official, unmodified Claude Code CLI you are logged into, so each turn counts against your Claude plan's limits; it drops Anthropic API keys from the CLI's environment and has no API-key fallback; Nous advises turning off extra usage in your Claude account if you do not want overage billing (as reported by Nous Research).

## Alternatives in the same category
- OpenClaw: https://agentgid.com/agents/openclaw/agent.md
- DeepSeek Harness: https://agentgid.com/agents/deepseek-harness/agent.md
- Manus: https://agentgid.com/agents/manus/agent.md
- Cloudflare OS: https://agentgid.com/agents/cloudflare-os/agent.md
- Muse: https://agentgid.com/agents/muse/agent.md
- Andee: https://agentgid.com/agents/andee/agent.md

## Links
- AgentGid page: https://agentgid.com/agents/hermes-agent/
- Source: https://github.com/NousResearch/hermes-agent
- Source: https://raw.githubusercontent.com/NousResearch/hermes-agent/main/README.md
- Source: https://en.wikipedia.org/wiki/Hermes_Agent
- Source: https://pypi.org/pypi/hermes-agent/json
- Source: https://formulae.brew.sh/api/formula/hermes-agent.json
- Source: https://hub.docker.com/v2/repositories/nousresearch/hermes-agent/

---
AgentGid's catalog is curated: every entry was checked against the maker's official pages (registry, repository, license and pricing pages) on the date shown; a license value shows the quote it stands on wherever AgentGid has one. Usage numbers come from AgentGid's own scan of public repositories set up for AI coding agents. Claims by makers are reported, not tested by AgentGid.
