TL;DR
Jack Roberts covers four major updates to the Hermes AI agent platform: a Ministry of Experts mixture-of-agents framework, computer control capability, a /learn skill system, and voice interaction. These updates collectively make Hermes more powerful, more accurate, and cheaper to run than using a single model alone.
Key points
1
Ministry of Experts runs multiple models (e.g. Opus 4.8, DeepSeek V4 Pro, GLM 5.2, GPT-5.5) in parallel, with an orchestrator synthesizing their answers — claiming 8% performance improvement over Opus 4.8 alone at a fraction of the cost via prompt caching
2
Hermes can now control your computer directly, handling apps and browser interactions that lack an API or MCP, running in the background across macOS, Windows, and Linux
3
The /learn command lets Hermes ingest any URL, folder, conversation, or pasted text and permanently store it as a reusable skill, so you never have to re-explain a process
4
Voice interaction is now supported via Whisper or ElevenLabs integration, allowing real-time spoken conversation with Hermes running continuously in the background
5
Additional quickfire updates include iMessage and WhatsApp integration, more reliable memory management, background agents for long tasks, a native desktop app with profiles, pets, and cron-free automation blueprints
Actionable insights
→
Set up Ministry of Experts using Opus 4.8 as orchestrator with DeepSeek V4 Pro and GLM 5.2 as reference models — add GPT-5.5 via a $20/month ChatGPT subscription to cut costs further while maintaining quality
→
Use an OpenRouter API key to connect Hermes to all major models in one integration, simplifying multi-model setup considerably
→
After any complex workflow or successful integration, immediately run /learn on that conversation so Hermes builds a permanent skill — eliminates future re-explanation and saves significant time
Notable quotes
“It doesn't even matter what the best model is. Why do I say that? Because if you really think about it, this list updates dynamically.”
“I did not invent the memory structure. I synthesized it.”
“It is fractionally more expensive for what they claim is actually better quality than just using 4.8 by itself.”
Worth watching?
⏭️
Worth watching the full video?
The key details on all four updates are fully captured here — only watch if you want the live screen walkthrough of setting up the Ministry of Experts panel inside the Hermes interface.
Saved you some time? The creator still deserves a like.
Found this by searching? That's the hard way.
summree watches the AI channels you follow and sends you the summary the moment a video drops. Nothing gets past you. From £4/month.
Try it free