Hermes Agent vs ChatGPT Dots: Control vs Convenience

Hermes Agent and ChatGPT Dots both offer an assistant that keeps working when you close the laptop, but they sit at opposite ends of one trade-off. Hermes is open-source software you run yourself: your machine, your model choice, your rules. Dots are managed agents from OpenAI: their cloud computers, their app ecosystem, your instructions. This comparison helps you decide whether control or convenience matters more for your business.

For the full background on each, see our Hermes Agent explainer and our ChatGPT Dots explainer. For the security practices that apply to both paths, read our small-business agent security guide.

Key takeaways

• Hermes Agent is free MIT-licensed software from Nous Research that you host, connecting to 20+ platforms with your choice of model provider.
• ChatGPT Dots, announced at OpenAI Dev Day on September 29, 2026 and powered by GPT-6 Astra, give each agent its own cloud computer and browser.
• Dots connect to more than 4,000 apps and are reachable through ChatGPT, mobile, Slack, and Microsoft Teams, with voice calls and planned SMS.
• Hermes keeps data on your equipment; Dots run on OpenAI infrastructure with isolated virtual machines and credential handling that hides passwords from the model.
• Both use approval tiers: Hermes through your own configuration, Dots through allowed, blocked, or approval-required actions plus Activity View supervision.
• Privacy-minded buyers with technical support lean Hermes; time-poor buyers already on ChatGPT plans lean Dots or a managed AI internal assistant.

The trade-off in one paragraph

Choose Hermes Agent when the deciding factors are data control, tool independence, and freedom from any vendor's plan: you run the software, pick from Nous Portal's 300-plus models, OpenRouter, OpenAI, or your own endpoint, and nobody can change the terms underneath you. Choose ChatGPT Dots when the deciding factors are speed of setup, breadth of ready-made app connections, and having OpenAI operate the computers, approvals, and safeguards. The rest of this article fills in what each side of that paragraph costs you in money, effort, and daily habit.

What each one is

Hermes Agent is an open-source assistant with the tagline "the agent that grows with you." It runs on your computer or server under the MIT license, with no paid tier and no hosted version. One gateway connects it to more than 20 platforms, including Telegram, Discord, Slack, WhatsApp, Signal, Microsoft Teams, Google Chat, email, and SMS. Its signature is the learning loop: agent-curated memory, autonomous skill creation, self-improving skills, and search across past sessions. Around that core sit a built-in cron scheduler, isolated subagents for parallel work, grounded research with citations and a fact-checking mode, real-time voice, and integration plumbing including the A2A v1.0 agent protocol, signed outbound webhooks, and seven terminal backends.

ChatGPT Dots are OpenAI's always-on agents, announced at Dev Day on September 29, 2026. Each Dot gets its own cloud computer and browser: an isolated virtual machine on OpenAI's infrastructure rather than software on your laptop. Dots connect to more than 4,000 apps through OpenAI's plugin ecosystem and work across ChatGPT, mobile, Slack, and Microsoft Teams, with voice calls supported and SMS planned. They do proactive work with guardrails: background research runs read-only, actions can be allowed, blocked, or set to require approval, and an Activity View lets you watch a Dot work and redirect it. Sensitive tasks such as changing a password stay with the user, credentials can be used without exposing passwords to the model, and Dots can be stopped if systems detect malicious instructions. For businesses, OpenAI is testing specialist Dots for procurement, invoice processing, customer support, and commercial contracts, and is working with Microsoft to bring them under Microsoft Agent 365 security controls.

Head-to-head comparison

One caution on freshness: Dots were announced September 29, 2026, the day before this article, so rollout details will move. Confirm current availability for your plan and region before committing. Hermes, released in early 2026 with v0.20.0 Herald in August and v0.21.x in September, is the more settled quantity.

Data and governance: who holds what

With Hermes, governance is yours to build. Data rests on your equipment and with the model provider you select. Sandboxing, credentials, approvals, logging, and updates are your controls to configure. The learning loop adds one job unique to Hermes: reviewing what the agent remembers and which skills it created, correcting wrong preferences early.

With Dots, governance is a mix of OpenAI's platform and your settings. Isolation comes from the per-Dot virtual machine. Credential design keeps passwords away from the model. Malicious-instruction detection can stop a compromised Dot. Your part is configuring action tiers thoughtfully, watching Activity View, and keeping sensitive tasks with humans. Neither side is automatically safer. Judge by which set of controls your business will actually maintain: self-built controls you review weekly, or platform controls you supervise daily.

Channels and the daily feel

Day to day, Hermes feels like texting a capable colleague. Because the gateway reaches the apps each person already checks, adoption rarely requires behavior change: the technician texts, the office manager uses Slack or email, and the owner dictates from the phone.

Dots feel like delegating inside the ChatGPT universe. If your team already works in ChatGPT, Slack, or Teams, a Dot slots into existing habits with the advantage of 4,000 ready app connections and specialist business Dots on the way. If your team coordinates over WhatsApp, SMS, or Telegram groups, Hermes reaches them where Dots currently do not. Map your team's three most-used apps against the two channel lists before anything else; the tool that meets people where they are wins on adoption alone.

Cost and effort

Hermes costs model usage plus hosting plus setup help, with no software fee and no plan gate. A light deployment is inexpensive; the real budget line is the day or two of technical setup and the weekly review habit. Dots ride on ChatGPT plans you may already hold: Pro and Business Premium in eligible markets, first Dot included with extended limits in the first month, and Dot usage initially not counting toward plan limits.

Effort follows the same split as everything else. Hermes asks for setup and maintenance in exchange for independence. Dots ask for configuration and supervision in exchange for speed.

Verdict: privacy-minded vs time-poor

Whichever you lean toward, pilot one workflow for two weeks before expanding: a briefing, an invoice triage, or a research watch. Approvals stay with a human on money, messages, and credentials either way.

FAQ

What is the core trade-off between Hermes Agent and ChatGPT Dots?

Hermes Agent is free, open-source software you run on your own equipment with the model provider of your choice, so you control data, tools, and updates. ChatGPT Dots, announced September 29, 2026, are managed always-on agents where each Dot gets its own cloud computer and browser, connects to more than 4,000 apps, and is reached through ChatGPT, mobile, Slack, or Teams. Control versus convenience is the whole decision.

Where does my business data live with each option?

With Hermes, data stays on your machine or server plus whichever model provider you choose, such as Nous Portal, OpenRouter, OpenAI, or your own endpoint. With Dots, work happens on OpenAI-operated cloud computers in isolated virtual machines, with credentials usable without exposing passwords to the model. Privacy-minded buyers who want data at home lean Hermes; buyers comfortable with OpenAI infrastructure lean Dots.

Which channels do Hermes Agent and ChatGPT Dots work in?

Hermes connects through one gateway to more than 20 platforms including Telegram, Discord, Slack, WhatsApp, Signal, Teams, Google Chat, email, and SMS, plus voice with barge-in. Dots work across ChatGPT, mobile, Slack, and Microsoft Teams, support voice calls, and have SMS planned. If your team lives outside the ChatGPT and Slack orbit, Hermes reaches further.

Which should a time-poor owner pick?

Usually ChatGPT Dots, if you already hold a qualifying plan. Dots roll out to ChatGPT Pro and Business Premium in eligible markets with the first Dot included, and managed approvals, Activity View supervision, and specialist business Dots reduce setup work. Hermes pays off when control matters more than speed of setup and someone will own the server.

Next step

Name the workflow you would delegate first and check it against the verdict table: where your team talks, where your data must live, and who will supervise. Then make it concrete with a prioritized plan: start your free six-step AI automation plan. If the choice still feels close, book a call and we will talk through it against your actual stack.