Hermes vs OpenClaw: which open-source AI agent fits your business?

Hermes vs OpenClaw: that is the comparison you make when you want to run an open-source AI agent yourself. Both agents run on hardware you choose. You talk to them through chat channels you already use. They remember what you tell them and work with the language model you pick. The differences are in how the agent learns, how you limit permissions and how you switch from one to the other.

In this article, we put Hermes and OpenClaw side by side on the points that matter to a business. Every fact comes from the official repository or documentation. You will also read what the warning from the Dutch Data Protection Authority means and how to reduce the risks. Voltti installs Hermes for businesses in Belgium and the Netherlands. We are open about why, without knocking OpenClaw.

A black capsule shape and a white faceted column on a dark plinth, connected by one thin cobalt-blue line
Two open-source agents with a lot in common. The differences are in the details.

What is OpenClaw?

OpenClaw is an open-source AI assistant that runs on your own computer. You reach it through channels such as Discord, iMessage, Slack, Teams, Telegram and WhatsApp, and through apps for macOS, iOS, Android, Windows and Linux. That is how the README on GitHub describes it. At its core is the Gateway: one process that manages sessions, tools and the connections to your channels. The OpenClaw documentation calls it a self-hosted gateway between your chat apps and AI agents.

Peter Steinberger started the project. It is now stewarded by the OpenClaw Foundation, an independent nonprofit. The code is under the MIT license, and there is no paid tier or hosted service. You may still come across the project under an older name, such as Clawdbot.

What is Hermes?

Hermes Agent is an open-source AI agent by Nous Research, also under the MIT license. According to the Hermes documentation, the agent creates skills from experience, improves them during use and builds a picture of who you are across sessions. You work with it in the terminal or in the desktop app. Through one gateway, you can also reach the agent on more than twenty chat platforms, such as Telegram, Slack, WhatsApp, Signal, Microsoft Teams and email.

For a broader explanation, read Hermes Agent explained. How Voltti uses Hermes for businesses is described on the Hermes AI agent page.

How are Hermes and OpenClaw alike?

More than you might expect. Both are free and open source under the same license. You choose where the agent runs. You can use a cloud model through an API key or a subscription, or a local model. Both keep long-term memory in plain text files, including MEMORY.md and USER.md. Both have skills, scheduled tasks and a gateway that serves several channels at once. And in chat channels, neither one simply responds to unknown senders by default: you pair or approve someone first.

Hermes vs OpenClaw: the differences at a glance

The table below summarizes the official repositories and documentation, checked on November 20, 2026. Both projects update often, so details can change.

TopicHermesOpenClaw

Maker and governance

Nous Research, MIT license

OpenClaw Foundation (nonprofit), MIT license

Channels

More than 20 platforms from one gateway, plus terminal and desktop app

More than 20 channels from one Gateway, plus desktop and mobile apps

Memory

MEMORY.md and USER.md, plus searchable past sessions

MEMORY.md, USER.md and daily notes, summarized in the background

Skills

The agent writes and improves its own skills; approval per change can be turned on

Learns skills from finished work, with versions and undo; a propose mode can be turned on

Models

API key, subscription (such as ChatGPT/Codex; Claude Max with extra usage credits) or local model

API key, subscription (such as ChatGPT/Codex or Claude through Claude CLI) or local model

Installation

Script for Linux, macOS and WSL2, native Windows and a desktop app

Script for macOS, Linux and WSL2 and for Windows; installs Node.js if needed

Commands

A helper model assesses the risk of dangerous commands; uncertain cases ask for your approval

Approval per command; without a sandbox, main-session tools run on the host

Switching

Imports an OpenClaw setup, with a preview first

Imports memory from Hermes (MEMORY.md and USER.md)

Which differences matter to a business?

The biggest difference is how the agent learns. Hermes writes a skill once it has worked out a way of doing something, and adjusts it later. By default, that happens freely. According to the skills documentation, you can have every change staged until you approve it. That also works from a chat channel. OpenClaw reviews finished work in the background and saves lessons as learned skills, with a history and an undo option. There, too, you can choose a mode that makes a proposal first.

For a team, what matters most is control. Which skills, plugins and integrations does the agent use, and who has reviewed them? Our advice is the same for both agents. Keep your team’s skills, plugins and MCP servers in a GitHub repository owned by your business. Deploy them from there to every installation. That way, the team works on the same way of working and you can trace every change.

The approach to commands differs. Hermes checks every command against a list of dangerous patterns. In the default mode, a helper model assesses the risk and an uncertain case asks for your approval. That is how the Hermes security documentation describes it. OpenClaw binds an approval to that exact command and denies by default when nobody is available to approve. The README does point out that main-session tools run on the computer itself unless you configure a sandbox. Both agents can also run commands in a container.

What does the Dutch Data Protection Authority say about OpenClaw?

On February 12, 2026, the Dutch Data Protection Authority (AP) warned about the security risks of AI agents such as OpenClaw. According to the AP, about one fifth of the plugins for OpenClaw contain malware, based on findings from security researchers. The AP also mentions hidden instructions in websites, emails and chat messages. On top of that, there are vulnerabilities that let an attacker take over a system remotely, and data leaks from an incorrect installation. The regulator calls on users and organizations not to use such agents on systems with privacy-sensitive or confidential data.

The warning names OpenClaw, but it covers OpenClaw and similar AI agents. So it applies to Hermes as well. Any agent that runs commands on its own and has access to your email or files carries the same kind of risk. The setup is what makes the difference.

How do you reduce those risks?

The measures below follow the AP’s advice and the security documentation of both projects. They apply to Hermes and to OpenClaw.

  • Keep the installation off the public internet. Put the agent behind a firewall and make the installation reachable only through the company network or a VPN such as Tailscale. That is how Voltti always installs Hermes. On a regular install, OpenClaw binds the Gateway to the local computer. The project also has a separate checklist for anyone who exposes it anyway.

  • Only let known people in. Use the pairing codes and sender allowlists that both agents have.

  • Give each system its own key with as few permissions as possible. Read-only where that is enough.

  • Be careful with third-party skills and plugins. Only install what you can read and whose source you know.

  • Have a person approve actions with major consequences first, and rotate keys if you think they have leaked.

Hidden instructions in emails or web pages are called prompt injection. Read more in prompt injection and API keys. For a practical list before installation, see our safety checklist for AI and business data. How we protect data and access is explained on the security and privacy page.

Can you switch from OpenClaw to Hermes?

Yes. Hermes has an official migration guide. The command hermes import-agent openclaw comes from an official plugin and always shows a full preview of what will be imported first. Then it asks for confirmation. It brings over the persona, memory, daily notes, skills and default model, among other things. The command only brings API keys along if you ask for them. If you start the import from hermes setup, they do come along after the preview and your confirmation. Settings without an equivalent, such as scheduled tasks, go into an archive so you can recreate them yourself. Before the import, Hermes makes a backup of its own configuration by default.

The other direction exists too. According to the OpenClaw memory documentation, OpenClaw can import MEMORY.md and USER.md from Hermes. Configuration, keys and skills are not part of that.

Why does Voltti work with Hermes?

OpenClaw is a full-fledged agent with a large community. We chose Hermes for a few concrete reasons. The agent turns recurring work into a skill on its own, and you can put that writing step behind an approval. You are free to choose the language model, through an API or a subscription. And separate profiles keep different environments apart. Our own Hermes installation at Voltti works this way: each environment has its own settings, its own memory and its own keys with minimal permissions.

You can install either agent yourself. How that works for Hermes is covered in installing Hermes Agent. If you are torn between doing it yourself and having it done, read your own AI assistant for your business. If you are comparing Hermes with a chat assistant instead, see Hermes vs ChatGPT and Copilot.

Where do you start?

Start with the work, not the tool. Pick one task that comes back every week and look at which systems the agent needs for it. In a pilot, you test that task alongside your current way of working, with the permissions the task needs and no more. A pilot usually takes 2 to 4 weeks, from the first conversation to the evaluation. How that works is described in our approach.

Portrait of Seppe Gadeyne

Seppe Gadeyne is a developer and the owner of Voltti. He sets up AI agents for companies in Belgium and the Netherlands.

Updated on

Frequently asked questions

01What is OpenClaw?

OpenClaw is an open-source AI assistant that runs on your own computer or server. You talk to it through chat channels such as WhatsApp, Telegram, Slack or Teams. The OpenClaw Foundation, an independent nonprofit, stewards the project. The code is under the MIT license.

02Is OpenClaw free?

The software is free and open source, and there is no paid tier. You do pay for the language model, unless you run a local model. And you pay for the computer or server the agent runs on. The same goes for Hermes.

03Can I switch from OpenClaw to Hermes?

Yes. Hermes has an official import feature for OpenClaw. It shows a preview first and then brings over the persona, memory, skills and default model, among other things. The command only brings API keys along if you ask for them. Settings without an equivalent, such as scheduled tasks, go into an archive for you to recreate.

04Which is safer, Hermes or OpenClaw?

That mostly depends on the setup. The warning from the Dutch Data Protection Authority covers OpenClaw and similar agents, so Hermes as well. Keep the installation off the public internet and only let known senders in. Give each system minimal permissions, be careful with third-party plugins and have important actions approved first.

05Can I use Hermes or OpenClaw with a local model?

Yes, both work with local models through your own model server, alongside cloud models through an API key or a subscription. A local model keeps your data on your own hardware. You do need a computer that is powerful enough for the model and the task.

Torn between Hermes and OpenClaw?

Discuss your pilot