Hermes Agent: What It Is, How to Install It and Automate Work (2026 Guide)


Hermes Agent is an open-source, self-hosted AI agent built by Nous Research and released under the MIT license. This guide covers what Hermes Agent is, how to install it, which model to connect, what it costs, and how to set up a first scheduled task without handing it too much access. Everything below comes from the official docs, linked in the Sources section.
Not to be confused with the Hermes language models Nous also trains. Hermes Agent is the agent software. You point it at a model provider of your choice.
What is Hermes Agent?
The docs describe it as a self-improving agent with a built-in learning loop. It keeps memory across sessions, creates skills from experience and improves them as it works. It is not tied to an IDE and not a wrapper around one API. It runs wherever you put it, and you talk to it from a terminal, the Hermes Desktop app, or chat apps.
The pieces that matter for everyday work:
- Memory. Agent-curated memory that carries across sessions, with cross-session recall.
- Skills. Reusable procedures the agent writes from tasks it has done. Skills follow the open agentskills.io standard, so they are portable.
- Scheduled tasks. Built-in cron that delivers results to the chat you already use.
- Messaging gateway. One background process that connects Hermes to Telegram, Discord, Slack, WhatsApp, Signal, email and more (the docs list 20+ platforms).
- Tools. 60+ built-in tools, plus MCP server support to add more.
- Subagents. Isolated helpers that work on parallel jobs.
- Terminal backends. Run it locally, in Docker, over SSH, or on Daytona, Singularity or Modal.
How to install Hermes Agent
- Pick an installer. On macOS or Windows, the docs recommend the Hermes Desktop installer. For command line only, Linux, macOS, WSL2 and Android (Termux) use a one-line install script from the docs. Native Windows has a PowerShell command. Run the exact command from the quickstart rather than copying it from a third-party page.
- Reload your shell. The docs tell you to run
source ~/.bashrcorsource ~/.zshrcafter the installer finishes. - Choose a provider. Run
hermes modelto pick interactively. The docs call this the single most important setup step.hermes setupoffers Quick Setup (Nous Portal), Full Setup (bring your own keys) and Blank Slate (almost everything switched off). - Check one real chat. If Hermes cannot complete a normal conversation, do not add features yet. The quickstart's rule of thumb is to get one clean chat working first.
- Add the gateway later. Once the CLI works, run
hermes gateway setupand connect one platform.
Which model does Hermes Agent use?
Hermes does not ship one fixed model. It works with any OpenAI-compatible API. The docs list Nous Portal, OpenRouter, OpenAI, Anthropic, GitHub Copilot, Google AI Studio, DeepSeek, Hugging Face, AWS Bedrock, local options like Ollama and LM Studio, and a custom endpoint for vLLM or SGLang.
One hard requirement: the model needs a context window of at least 64,000 tokens. Smaller windows are rejected at startup. Most hosted models meet that.
How to set up Hermes Agent scheduled tasks
You can schedule in plain language or with a command. From the cron docs:
- In chat:
/cron add "every 2h" "Check server status" - From the terminal:
hermes cron create "every 1h" "Summarize new feed items" - Or just ask: "Every morning at 9am, check Hacker News for AI news and send me a summary on Telegram."
Jobs can also run on a schedule with no model at all (a script whose output is delivered as is), or fire when a webhook arrives instead of on a timer.
A first automation worth trying: every weekday morning, summarize a few sources and send the result to your chat. That job reads and reports. It does not send email, change records or spend money. Run it for a week and read the output before you trust it with more.
Hermes Agent messaging gateway setup
Run hermes gateway setup for the interactive setup, then hermes gateway install to run it as a service and hermes gateway start to start it. By default the gateway denies everyone who is not on an allowlist or paired by direct message. The docs call this the safe default for a bot with terminal access. Keep it that way.
Hermes Agent safety, cost and limits
- Cost. Hermes itself is free. You pay your model provider, or a Nous Portal subscription if you choose that route. Local models are free to run.
- Dangerous commands. Hermes checks commands against a list of dangerous patterns and asks you before running a match. The default mode is
smart. For cron jobs the default is to deny a dangerous command rather than auto-approve it. - Access. Give it only the accounts and folders a job needs.
- Model changes. Per the cron docs, a job that is not pinned to a model uses whatever your main model is when it fires. Pin the model if you want stable output.
- Review. Keep a human step before anything goes to a customer or changes live data.
| Tool | What the docs describe |
|---|---|
| Hermes Agent | Memory and reusable skills, scheduled jobs and a messaging gateway. Terminal execution can use local, Docker, SSH and other backends. |
| Agent Zero | A Dockerized Linux desktop, browser and document editing, with project-scoped memory and optional access to host files. |
Sources: Hermes README · Agent Zero README
Is Hermes Agent right for you?
Hermes fits if you want one assistant that lives on your own server and reports to chat. If you would rather build a visual flow with fixed steps, see my post on Dify. If your work is writing code across several branches, look at Orca. If you want an agent with its own Linux desktop in Docker, see Agent Zero.
Want help picking or setting up a workflow? Send the task and the tools you already use through the contact page.
Frequently asked questions
What exactly is Hermes Agent?
Hermes Agent is an open-source, self-hosted AI agent from Nous Research, released under the MIT license. Per its docs it keeps memory across sessions, creates reusable skills from experience, runs scheduled tasks, and connects to chat apps such as Telegram, Discord, Slack and WhatsApp through a messaging gateway.
What agent does Hermes use?
Hermes Agent does not ship one fixed model. You connect a model provider and it works with any OpenAI-compatible API. The docs list Nous Portal, OpenRouter, OpenAI, Anthropic, Google AI Studio, DeepSeek, local options like Ollama and LM Studio, and custom endpoints. The model needs a context window of at least 64,000 tokens.
How expensive is Hermes Agent?
The software is free. The docs say Hermes itself is free and open source under the MIT license, and you pay only for the model usage from your chosen provider. Local models are free to run. A Nous Portal subscription is the option that bundles models and tools under one bill.
Is Hermes Agent free or paid?
Hermes Agent is free and open source (MIT). What can cost money is the model provider you connect, such as an API key from OpenRouter or OpenAI, or a Nous Portal subscription. Check usage during the first week.
How are people using Hermes Agent?
The docs focus on a few patterns: scheduled automations with results delivered to chat (for example checking server status every 2 hours or summarizing new feed items every hour), an always-on assistant reached through Telegram or Discord, and parallel work through subagents. Start with a low-risk read-and-report job.
What is required to run Hermes Agent?
You need to install it (the Hermes Desktop installer on macOS or Windows, a one-line install script on Linux, macOS, WSL2 or Android via Termux, or a PowerShell command on native Windows) and connect a model provider. The model must have a context window of at least 64,000 tokens. The docs say it can run on a small VPS, a GPU cluster or serverless infrastructure, so there is no fixed hardware requirement beyond what your model needs.
What is the difference between Hermes and Hermes Agent?
Nous Research is the lab behind the Hermes language models. Hermes Agent is separate software: an agent that runs on your machine or server and uses a model provider you choose. Pick a model for it; it is not itself the model.
Sources
Frequently asked questions
What exactly is Hermes Agent?
Hermes Agent is an open-source, self-hosted AI agent from Nous Research, released under the MIT license. Per its docs it keeps memory across sessions, creates reusable skills from experience, runs scheduled tasks, and connects to chat apps such as Telegram, Discord, Slack and WhatsApp through a messaging gateway.
What agent does Hermes use?
Hermes Agent does not ship one fixed model. You connect a model provider and it works with any OpenAI-compatible API. The docs list Nous Portal, OpenRouter, OpenAI, Anthropic, Google AI Studio, DeepSeek, local options like Ollama and LM Studio, and custom endpoints. The model needs a context window of at least 64,000 tokens.
How expensive is Hermes Agent?
The software is free. The docs say Hermes itself is free and open source under the MIT license, and you pay only for the model usage from your chosen provider. Local models are free to run. A Nous Portal subscription is the option that bundles models and tools under one bill.
Is Hermes Agent free or paid?
Hermes Agent is free and open source (MIT). What can cost money is the model provider you connect, such as an API key from OpenRouter or OpenAI, or a Nous Portal subscription. Check usage during the first week.
How are people using Hermes Agent?
The docs focus on a few patterns: scheduled automations with results delivered to chat (for example checking server status every 2 hours or summarizing new feed items every hour), an always-on assistant reached through Telegram or Discord, and parallel work through subagents. Start with a low-risk read-and-report job.
What is required to run Hermes Agent?
You need to install it (the Hermes Desktop installer on macOS or Windows, a one-line install script on Linux, macOS, WSL2 or Android via Termux, or a PowerShell command on native Windows) and connect a model provider. The model must have a context window of at least 64,000 tokens. The docs say it can run on a small VPS, a GPU cluster or serverless infrastructure, so there is no fixed hardware requirement beyond what your model needs.
What is the difference between Hermes and Hermes Agent?
Nous Research is the lab behind the Hermes language models. Hermes Agent is separate software: an agent that runs on your machine or server and uses a model provider you choose. Pick a model for it; it is not itself the model.
Sources
Ready to implement this for your brand?
Stop reading about growth and start engineering it. Our autonomous marketing systems and SXO strategies are battle-tested and ready to deploy.
Initiate Strategy Session


