Hermes Agent + Open WebUI Just Changed AI Agents Forever
The video demonstrates how to integrate Open Web UI with the Hermes AI agent, providing a modern chat interface as an alternative to Hermes's limited terminal UI. The setup is accomplished by having Hermes itself read the GitHub documentation and configure the Docker-based installation automatically. The result gives users a ChatGPT-style interface with file uploads, model switching, image generation, and mobile access.
Summary
The video is a practical walkthrough of setting up Open Web UI as a front-end chat interface for the Hermes AI agent. The presenter begins by explaining the motivation: while Hermes has its own web dashboard for managing scheduled tasks, skills, and sessions, it lacks a direct chat option, making it cumbersome for day-to-day interaction. Open Web UI solves this by providing a polished, ChatGPT-style messaging interface.
A notable aspect of the setup process is that Hermes itself is used to automate the installation. The presenter pastes GitHub repository information and documentation into Hermes, which then reads through it, installs dependencies, configures an API server and environment file, and deploys everything inside Docker — all without manual configuration steps from the user.
Once running, the presenter highlights several advantages of the Open Web UI integration over the native Hermes terminal or dashboard. These include the ability to attach files directly in the chat (rather than manually typing file paths), a code preview pane similar to ChatGPT's canvas, voice and video call support, mobile accessibility, web search and browsing capabilities, and image generation integration. The interface also allows users to switch between multiple models — including local Ollama models, DeepSeek, and external APIs — from a single workspace.
The presenter also demonstrates the 'new model' feature in Open Web UI's workspace, which lets users create customized agent profiles with specific system prompts, tools like web search or image generation, and uploaded knowledge bases. This effectively allows multiple specialized agents to coexist and be managed from the same interface.
The video concludes with a brief demonstration of Hermes building a ping pong game through the chat interface, previewing the output in-browser, and self-correcting errors when prompted. The presenter wraps up by directing viewers to the 'AI Profit Boardroom' community for a full guide, a 2-hour Hermes course, a 6-hour OpenClaw course, and a step-by-step tutorial on Hermes V0.11.
Key Insights
- The presenter points out that the official Hermes dashboard, while functional for managing tasks, skills, and sessions, does not include a direct chat option — which he identifies as a significant limitation that motivated the entire Open Web UI integration.
- Rather than manually configuring the setup, the presenter uses Hermes itself to read the GitHub documentation and automatically install, configure, and deploy Open Web UI inside Docker, demonstrating Hermes's ability to act as its own setup agent.
- The presenter argues that using the terminal UI is impractical for file-heavy workflows because users must manually type full file paths, whereas Open Web UI allows direct file, note, and knowledge attachment within the chat interface.
- The presenter explains that Open Web UI's workspace allows users to create multiple distinct agent profiles — each with its own system prompt, tool access (e.g., web search, image generation), and knowledge base — enabling different specialized versions of Hermes to coexist.
- The presenter acknowledges that the current Open Web UI workaround may be temporary, suggesting that the Hermes team will likely add a native chat feature to their official mission control dashboard in the future.
Topics
Full transcript available for MurmurCast members
Sign Up to Access