Hermes Agent

Hermes Agent is an open-source autonomous agent designed to run on your own server and keep learning as it goes. It is aimed at developers and technically inclined users who want a persistent agent that can do more than answer prompts — it can maintain memory, build skills, and carry work across sessions.

According to the documentation, you install it with a shell script and then configure it with hermes setup. It connects to multiple interfaces, including Telegram, Discord, Slack, WhatsApp, Signal, Email, and the CLI, and it also includes browser and web automation capabilities.

This is a more agentic system than a simple chat wrapper, since it can run unattended scheduled jobs, spawn isolated subagents, and use different sandbox backends such as local, Docker, SSH, Singularity, and Modal. The site also mentions web search, vision, image generation, text-to-speech, and multi-model reasoning, so it appears built for broader task execution rather than one narrow workflow.

A few details are still unclear from the page: pricing is not publicly listed, and the exact model providers behind its “multi-model reasoning” are not specified. The product is open source under the MIT License, which improves transparency, but it is self-hosted and therefore better suited to users comfortable managing their own infrastructure. If you want a fully managed no-code assistant, this is probably not the right fit.