Video: "Hermes Cloud is TOO EASY" by Julian Goldie on YouTube.
What Hermes Cloud actually is
Until Hermes Cloud, running a Hermes Agent meant running it locally. Your machine had to be on, your installation had to be open, and anything that needed to run overnight or over a weekend required leaving hardware switched on. Hermes Cloud removes that constraint. It deploys an agent instance to a cloud container that Hermes manages, so the agent keeps running regardless of what happens on your end.
It is not a stripped-down version of Hermes. The cloud instance has access to the same tool system — browser control, file access, API calls, terminal commands, memory — as the local installation. The only thing that changes is where the agent runs.
What the setup looks like
Julian walked through the deployment in the video and the process is short. You sign in to the Hermes Cloud portal, give the agent a name, choose a model from the available list, and deploy. The agent appears in your dashboard and starts running. There is no infrastructure to configure and no container management to do yourself.
The model list at launch runs to over 200 options. That includes the models already available in the local Hermes stack — Kimi K3, Sonnet 5, Opus, GPT-5.6, Grok-4.5 — plus others accessible through OpenRouter and similar aggregators. You pick the model when you deploy and can adjust it later.
What 24/7 actually means in practice
The useful thing about always-on deployment is not speed — Hermes agents are already fast. It is reliability over time. An agent that checks a news feed every hour, monitors competitor pricing daily, or runs a reporting job every Friday morning needs to be alive when those triggers fire. A local agent cannot guarantee that without someone managing it.
Hermes Cloud hands that problem off. You configure the jobs and the schedule in the Hermes skill system, deploy to cloud, and the tasks run. Julian's framing is that this is closer to hiring a member of staff than renting software: you set goals and leave it running, rather than giving it instructions each session and waiting for output.
How it fits the wider Hermes stack
Hermes has been adding components steadily. Oracle handles news monitoring. Apollo is the voice interface covered in the related article below. Astros tracks competitors. The local Hermes Agent handles task execution and skill routing. Hermes Cloud is the infrastructure layer that keeps all of it running without a local machine in the loop.
The pattern is modular: each component handles one job, and you combine them for what your situation requires. A business that wants a competitor monitoring agent running overnight, flagging changes each morning, would combine Astros (the monitoring logic) with Hermes Cloud (the always-on host). Neither is useful without the other; together they do something a local setup cannot.
What it still requires from you
Hermes Cloud does not remove the need to configure the agent correctly. You still define the skills, set the schedules, write the goals, and manage what the agent does with the output. The cloud layer handles execution reliability; it does not handle judgement.
There are also API costs to consider. Hermes Cloud hosts the agent, but model calls still go through whichever provider you have configured. Running a capable model continuously across multiple tasks will generate API spend. Julian does not cover pricing in the video, so treat that as an open question before deploying anything resource-intensive.
Where this connects to NordSys
If you want an AI agent running inside your business — configured with your goals, handling tasks on a schedule, and reporting back without needing constant instruction — that is what our AI Agents service provides. We configure it, name it, and manage it for you, without the setup overhead. From £6 a day.
See our AI Agents →