Big News: Hermes Gets a New Chat UI + Affordable $9.99 Plan with Unlimited Tokens!
We’ve been busy at Agent37 and have some exciting updates to share about Hermes Hosting!
1. Chat UI for Non-Technical Users – HermesWebUI
When we first launched Hermes, it was primarily terminal-based, which worked well for advanced users but wasn’t the most accessible for non-technical folks. We received a ton of feedback asking for an easier way to interact with Hermes.
So, we listened! HermesWebUI is a new chat interface that provides a simple and user-friendly experience, making Hermes easy to use for everyone! No technical skills needed just chat and get things done.
Why it matters: The new chat UI removes the technical barrier and brings Hermes to a wider audience, making it easier to use for both beginners and experienced users.
Built using Python and vanilla JavaScript, HermesWebUI offers a lightweight solution with no frameworks or complex dependencies.
2. Introducing the $9.99 Plan with Unlimited Tokens
Another frequently requested feature is our affordable token plan. We’ve rolled out a $9.99/month plan with unlimited tokens for fast Open Source models, like Qwen 3.5 Flash.
Why it matters: Many of our users wanted to test Hermes and OpenClaw without breaking the bank on expensive tokens. Now, with this new plan, you can explore all the powerful models without worrying about running out of credits!
Why These Updates Matter:
HermesWebUI provides a much-needed solution for users who are finding it difficult to interact with Hermes through a terminal. Now, anyone can use it!
The $9.99 plan is a game-changer for those who need an affordable option to try out powerful models like Qwen 3.5 Flash without spending too much.
What do you think of the new changes? Would love to hear your thoughts and any feedback on the new features!
The "Python and vanilla JavaScript, no frameworks" line is the part I most want to underline here. As someone shipping a small indie iOS app solo, the lightweight stack choice keeps paying back every week — bundle size, debugging speed, and zero dependency-update churn. I've watched friends rebuild three times because their early framework choice locked them into upgrade treadmills. The 9.99 unlimited-token plan is also a brave pricing call — curious how you're guarding against power-user abuse without metering: rate limiting per-IP, request size caps, soft caps with email warnings? And do non-technical users actually feel the value of "unlimited," or do they default-assume they'd never use that much anyway?
Yeah we did look into managed services early on. For this we wanted something that just works out of the box without adding extra complexity. Trying to keep it simple for people who don’t want to deal with setup at all
Useful info. Have you tried managed services?
That’s always the tough part with unlimited plans. We are trying to keep things efficient on the backend so it doesn’t get out of hand, while still letting users feel like they are not being restricted all the time.
That’s a smart approach users like flexibility but keeping the system stable matters too.
The $9.99 unlimited tokens angle is smart — pricing anxiety is one of the biggest reasons people don't stick with AI tools. Curious how you're handling the cost side as usage scales up? That's always the tricky part with unlimited plans. Following this one — congrats on the launch
Thankyou very much. Well biggest challenge so far has been making it simple without taking away flexibility. It’s easy to build for one type of user, but making it work for both beginners and more technical people takes a bit more thought.
That's very interesting, having AI in charge is a whole different ballgame... I hope to try it out, because I'll definitely need this kind of tool. What's been your biggest challenge so far?
Thankyou that means a lot. Definitely give it a try when you get a chance. Would be interesting to hear how it fits into your workflow.
I'm Dex. Full-stack developer specializing in LLM integrations — self-hosted, cost-efficient, privacy-first setups. Been building in public for 2 years alongside my startup work. Projects include an open-source agent framework and local LLM tooling. Available for freelance — React, Node.js, Python, AI/LLM integrations. If you're building something in this space and need a dev, DM me.
Thank you for sharing this, Dex. It's always nice to connect with people who are developing in the LLM field. There may be some overlap because we're currently investigating integrations and use cases related to self-hosted + economical solutions.
Hey Vishnu — thanks for the reply. Agent37 looks genuinely interesting. The zero-setup angle is smart; most people building in this space underestimate how much friction infra adds before anyone even gets to the actual agent logic.
I'd love to explore the overlap you mentioned. My angle on self-hosted + economical is mostly around making local models production-usable on constrained hardware — I've been running agent loops on 6GB VRAM and writing about where it breaks. That's probably adjacent to what you're investigating.
A few quick things I can offer depending on what you need:
What does the overlap look like on your end? Happy to jump on a quick call or just keep it async — whatever works for you.
You can check out my work on GitHub (workspace-dex/open-agent)
The $9.99 plan is quite reasonably priced. I would like to know if this provides unrestricted access to all models or if there are limitations based on task complexity.
The $9.99 plan does give you access to most models but there are some high end models that might require a premium plan for more intensive use. But for everyday tasks and testing, this plan should cover all your needs.
This is actually a really strong update. The shift from terminal-based usage to a chat UI will make a big difference for non-technical users. The $9.99 unlimited token plan also feels like a smart entry point for testing and experimentation without worrying about limits.
Reminds me how some simple tools become popular just because they improve usability and reduce friction for users. Even in other niches, like content browsing and structured listings, platforms such as swigmenux show how clarity in presentation can improve user experience.
Overall, this looks like a solid step forward for Hermes.
We are undoubtedly going in that direction. We intend to increase model support and progressively add more sophisticated features while maintaining an easy-to-use chat interface. We are actively pursuing this balance.
Do you see this chat interface expanding to cover more advanced features or other AI models down the line? It would be cool to see how you’ll build on this!
This is just the beginning. We have a roadmap for adding more advanced features and AI models as we grow. We are excited to keep improving and offering even more functionality for all types of users
The pricing is great!But I am curious how do the tokens work with OpenClaw?Is it limited by the model you use,or does it stay consistent across all open source models?
Glad you like the pricing, Leena! With OpenClaw, the token usage depends on the complexity of the model you're using. Some models might consume more tokens due to their size and computational needs, but for most use cases, it’s pretty consistent
I appreciate how you're making it simpler for everyone to begin using the unlimited token plan. Startups and small developers would undoubtedly benefit from this. Do you think this pricing is long-term plan to draw in more customers?
Absolutely.. ee believe in making powerful tools accessible to smaller teams and individuals, and this pricing is part of our long-term strategy to keep things affordable. We want as many people as possible to experiment with Hermes without worrying about costs
The new UI and pricing plan are a huge win! I’m curious, what’s next for Hermes? Any plans to add more customization options or other cool features to the chat UI in the future?
Thankyou. We have some exciting plans in the works. Customization is definitely on the list, as well as additional features like more model integration. We’re focusing on making the experience even more tailored to users' needs..
How do you feel this new chat UI stacks up against what other platforms are doing? Any features you have taken inspiration from or is this all new ground?
While there are some tools out there with similar features, we’ve tried to make Hermes as intuitive and user-friendly as possible. We’ve taken inspiration from the best, but we’ve also added unique elements to suit our community’s needs
The chat UI looks great, however, I’m wondering how you’ll make sure the more experienced users don’t feel constrained by the new interface. Is it possible to alternate between the old and new modes?
Right now, the new interface is designed to be a clean, simple option for those who prefer it. For experienced users, we are considering more advanced settings and features in future updates to ensure everyone can still access the depth they need
I love that you are making it easier for non-technical users to get started. Will there be any kind of guide or tutorial available to help current users transition from the terminal interference to the new UI?
We’re working on a tutorial for existing users to help with the transition. It’ll cover everything from basic navigation to more advanced functionalities, so no one feels lost when switching to the new UI
I’m really excited about the unlimited tokens for $9.99. Does the plan include access to all the models available, or are some advanced models still off limits for this price point?
The $9.99 plan includes access to many of our popular models. Some of the more advanced models, especially those with higher computational requirements, may need a higher-tier plan, but we’ve made sure this plan is a solid starting point
The new Hermes Web UI is definitely a step in the right direction for user accessibility. But how does it compare in terms of speed and performance to the traditional terminal interface? Does it have the same capabilities?
The new UI is optimized for speed and efficiency, but for power users, the terminal interface will still offer some performance advantages for heavy tasks. We're working to make the chat UI as fast as possible, though, and more performance updates are coming soon
The need for improved UX is truly highlighted by this latest version. It is a victory to make Hermes usable by non-technical users without sacrificing any functionality.
Thanks we are definitely looking into ways to provide more customization options down the line. It’s all about balancing ease of use with personalization, so we’ll be sure to keep that in mind