Always-on AI agents
Hermes Agent vs OpenAI Dots
Both are agents that keep working when you close the laptop. One comes with an eligible ChatGPT plan, including Pro at $200/month, and OpenAI’s own model. The other is open source and runs on the model and chat app you pick. Here is how to choose.
Our recommendation: If you already pay for ChatGPT Pro or Business Premium and live in Slack or Teams, try your included dot first. Choose Hermes Agent if you want to pick the model, talk to the agent in Telegram, WhatsApp or Discord, keep costs low, or you are a Pro user in the EU, UK or Switzerland.
How we compared: We run OpenClaw Launch, which hosts Hermes Agent. This page compares published information; we have not tested Dots hands-on. Dots details were checked against launch coverage on September 29, 2026, the day of DevDay. Early details change quickly, so check OpenAI’s current terms before you buy.
What is OpenAI Dots?
OpenAI announced Dots at DevDay on September 29, 2026. A dot is a personal agent powered by GPT-6 Astra. It runs on its own cloud computer rather than your device, connects to more than 4,000 apps through OpenAI’s plugin ecosystem, and can work toward your goals 24/7. You can give it a name and talk to it in ChatGPT, Slack or Microsoft Teams, or hop on a voice call. Source: 9to5Google.
OpenAI says dots follow the same permissions and controls as ChatGPT, and you can set rules for when a dot may act on its own and when it must ask you first. Text message support was announced as coming soon rather than available at launch. OpenAI also describes future “specialist” dots and teams of dots; today one dot is included. Source: TechCrunch.
What is Hermes Agent?
Hermes Agent is an open-source agent from Nous Research. It keeps memory across sessions, creates reusable skills from work it has done, runs scheduled jobs and talks to you through messaging apps. It works with the model of your choice. Source: Hermes Agent on GitHub.
You can run Hermes on your own server or deploy it on OpenClaw Launch, where we manage the container and you connect Telegram, Discord, Slack, WhatsApp, WeChat or Feishu from the dashboard. Hermes can be self-hosted; Dots cannot.
Quick comparison
| Category | OpenAI Dots | Hermes Agent |
|---|---|---|
| Product | OpenAI’s always-on personal agent inside ChatGPT | Open-source agent from Nous Research, self-hosted or managed |
| Model | GPT-6 Astra | Your choice of supported model and provider, including your own API key |
| Where you talk to it | ChatGPT, Slack, Microsoft Teams and voice calls; text messages announced as coming soon | Telegram, Discord, Slack, WhatsApp, WeChat, Feishu and web chat on OpenClaw Launch; more adapters upstream |
| Runtime | A dedicated OpenAI-hosted cloud computer per dot | A managed container on OpenClaw Launch, or any machine you run it on |
| Tools | Plugins connecting to 4,000+ apps | Browser, terminal, files, MCP servers and skills |
| Background work | Works toward goals 24/7, with rules for when it may act without approval | Built-in cron scheduler that delivers results to your chat channel |
| How many agents | One dot included; more planned later | One instance on Lite, up to 3 on Pro |
| Who can get it | ChatGPT Pro ($200/month), Pro 500 ($500/month) and Business Premium; Enterprise with admin approval. Not for Pro users in the EEA, Switzerland or UK at launch | 30-minute free trial, then plans from $3 for the first month; available wherever OpenClaw Launch is |
| Self-hosting | No self-hosted version announced; runs on OpenAI’s cloud | Yes, the upstream project is open source |
Sources: TechCrunch, 9to5Google, The Next Web, Hermes Agent, and our pricing and integrations pages.
Model choice: GPT-6 Astra or your pick
Dots runs on GPT-6 Astra. If that is the model you want, Dots gives it to you without managing any model settings. The launch coverage does not mention any way to choose another model or bring your own key.
Hermes lets you switch models whenever you like. You can run a cheap, fast model for a daily digest and a stronger one for coding, or use your own provider key and pay the provider directly. OpenAI models are among the options; see our GPT-6.1 Sol guide for Hermes and the model catalog. More choice also means more to check: make sure the model you choose calls tools reliably before you trust it with scheduled work.
Where you talk to your agent
Dots is built around work tools: ChatGPT, Slack and Microsoft Teams, plus voice calls. That suits a company that already runs on Slack or Teams. Text messages are on OpenAI’s roadmap but were not live at launch.
Hermes covers more personal messaging apps. On OpenClaw Launch you can connect Telegram, Discord, Slack, WhatsApp, WeChat and Feishu, or use web chat. Upstream Hermes also ships adapters for Microsoft Teams, Signal, email and SMS through Twilio, which you configure yourself with your own accounts.
Background work and permissions
Both agents are meant to work while you are away. A dot pursues goals on its own computer and can check in with you. Hermes uses a built-in scheduler: you describe a recurring job in plain language, such as a morning summary of your inbox and calendar, and it delivers the result to your chat. See scheduling AI agent tasks for examples.
Dots has an explicit approval system: you decide which actions it may take alone and which need your sign-off. That is a real strength for work with consequences, such as sending email or merging code. Hermes can ask before running commands, and on OpenClaw Launch each agent runs in its own container. We do not claim this matches OpenAI’s credential handling or approval rules; see our security page for how we store keys. Whatever you pick, decide what the agent may read, change and send before you connect important accounts.
Pricing: part of a ChatGPT plan, or hosting plus AI usage
Dots has no standalone price. One dot comes with ChatGPT Pro ($200/month), the new Pro 500 plan ($500/month) and Business Premium, and Enterprise admins can turn it on. Talking to your dot does not use up your ChatGPT limits, but tasks it starts in Codex or ChatGPT Work do. OpenAI is also halving the Work and Codex usage in Pro 200 from October 30, 2026. Source: The Next Web.
- OpenClaw Launch trial: 30 minutes of web chat, no credit card required.
- Lite: $3 for the first month, then $6/month, with $1/month in AI credits.
- Pro: $20/month, with $10/month in AI credits and up to 3 instances.
Hosting and AI usage are separate budgets on OpenClaw Launch. Heavy use beyond the included credits needs a top-up or your own provider key. If you already pay for ChatGPT Pro for other reasons, the dot costs nothing extra. If you would only upgrade to get a dot, compare that $200/month with a Hermes plan plus the model usage your tasks actually need.
Which should you choose?
Choose OpenAI Dots if you are already on ChatGPT Pro or Business Premium, your team works in Slack or Teams, and you want GPT-6 Astra with OpenAI’s approval controls, app plugins and hosted computer.
Choose Hermes Agent if you want to pick the model, talk to your agent in Telegram, WhatsApp, Discord or WeChat, keep a small budget, self-host later, or you are a Pro user in a region where Dots is not available.
A good test for either one is a single recurring task with a clear result, for example three relevant updates every morning with a one-line reason for each. Run it for a week and look at missed runs, noise and how easy it is to correct. That tells you more than a launch demo.
Frequently asked questions
What is OpenAI Dots?
Dots is a personal agent OpenAI announced at DevDay on September 29, 2026. Each dot runs on GPT-6 Astra with its own cloud computer, connects to more than 4,000 apps through plugins, and works toward goals in the background. You reach it through ChatGPT, Slack, Microsoft Teams or a voice call.
Is Hermes Agent an alternative to OpenAI Dots?
For many jobs, yes. Both are agents that keep working in the background and report back in the chat apps you use. The difference is who makes the choices. Dots comes with OpenAI’s model and hosted computer as part of a ChatGPT plan. Hermes lets you pick the model, the chat channel and where it runs, and you take responsibility for that setup.
How much does OpenAI Dots cost?
OpenAI includes one dot with ChatGPT Pro at $200/month, the new Pro 500 plan at $500/month and Business Premium, and Enterprise workspaces can enable it. There is no standalone price. Chatting with your dot does not count against ChatGPT limits, but work it starts in Codex or ChatGPT Work does. Hermes on OpenClaw Launch costs $3 for the first month on Lite, then $6/month, with Pro at $20/month.
Can I use Dots in Europe or the UK?
Not on a Pro plan at launch. OpenAI said Dots will not reach Pro users in the European Economic Area, Switzerland or the UK for now. Hermes Agent is open-source software you can run anywhere, and OpenClaw Launch does not exclude those regions; your model provider’s terms still apply.
Can Hermes use GPT models like Dots does?
Hermes can call OpenAI models through a supported provider or your own API key, alongside Claude, Gemini, DeepSeek, Qwen and others. Using a GPT model inside Hermes is still a different product from Dots: you do not get OpenAI’s hosted computer, plugin catalog or approval system.
Related comparisons and guides
- Hermes Agent vs Manus Cue: another new personal-agent app
- Meta Muse vs OpenClaw Launch: Meta’s personal agent
- ChatGPT Operator vs OpenClaw Launch: OpenAI’s earlier browser agent
- Hermes Agent memory: how Hermes remembers context