Hermes Agent V0.12 Just Changed AI Agents Forever!
Hermes Agent V0.12, dubbed the 'Curator Release,' introduces an autonomous background agent that self-manages AI skills every 7 days, along with Spotify and Google Meet integrations. The update delivers a 57% faster startup time, support for 19 messaging platforms, and was built by 213 community contributors with over 1,096 code changes. Users can update with a single terminal command: 'hermes update.'
Summary
Hermes Agent V0.12, released on April 30th and called the 'Curator Release,' is presented as a major overhaul of the free, open-source AI agent tool. The presenter walks through what's new, how to install the update, and why these changes represent a significant leap in autonomous AI agent capability.
The headline feature is the Autonomous Curator — a background process that runs every 7 days automatically via the 'hermes curator' command. Before this update, Hermes would accumulate saved skills over time with no cleanup mechanism, leading to clutter, slower performance, and wasted tokens. The Curator addresses this by reviewing all saved skills, deleting outdated ones, merging similar ones, and generating reports on what it changed. Users can check its status at any time with 'hermes curator status,' making the system self-maintaining without user intervention.
Performance improvements are also notable, with startup time reduced by 57%, achieved by loading only necessary components on launch. The presenter contrasts this favorably with OpenClaw, which they note has become slower over time due to unnecessary plugin loading.
New integrations include Spotify, which allows users to play music, search songs, and manage playlists directly through the agent via a secure OAuth-style login wizard. Google Meet integration is another major addition, enabling the agent to join calls, enable captions, transcribe conversations, speak during meetings, and follow up afterward — meaning users can optionally send their agent to a meeting in their place.
Messaging platform support has expanded to 19 platforms total, with Microsoft Teams (platform 19) and Tencent Yunba (platform 18) being newly added. Four new AI providers are now supported: GMI Cloud, Azure, MiniMax, and Tencent. LM Studio integration allows users to run local AI models directly with Hermes, and the presenter demonstrates browsing available models including DeepSeek V4, Gemma-4, and others. ComfyUI and TouchDesigner are bundled as optional add-ons for AI image/video generation and visual creative coding respectively.
A remote model catalog feature means new AI models appearing on OpenRouter are automatically available in Hermes without requiring a Hermes update itself. The release involved 213 community contributors, 360 bug fixes, 1,096 commits, and 550 pull requests merged. The presenter concludes by directing viewers to the 'AI Profit Boardroom' community for a full setup guide and 30-day roadmap.
Key Insights
- The presenter explains that before V0.12, Hermes had no cleanup mechanism for saved skills, causing old skills to pile up, slow the system down, and waste tokens — a problem the new Curator agent directly solves by running autonomously every 7 days.
- The presenter claims Hermes Agent now starts 57% faster because it loads only what it needs when it needs it, and contrasts this with OpenClaw, which has gotten slower due to loading unnecessary plugins.
- The Google Meet integration allows the Hermes agent to join calls, transcribe them, and follow up afterward — with the presenter suggesting users could skip a meeting entirely and let their agent attend in their place.
- The presenter reveals a remote model catalog feature that allows new AI models released on OpenRouter to become available in Hermes automatically, without requiring users to update the Hermes software itself.
- The presenter states the V0.12 release was built by 213 community contributors and included 1,096 code commits, 360 bug fixes, and 550 merged pull requests, framing it as a community-driven open-source effort rather than a minor patch.
Topics
Transcript
[0:00] Hermes Agent V 0.12 just dropped. It's a free new update and it just got about 100 times more powerful. Your AI agent can now clean itself, think for itself, and run meetings for you all without you touching anything. It starts faster, it gets smarter every week on its own, it joins your Google Meets, transcribes everything, and follows up after. Even if you miss a call, your agent can go instead of you. It can even control your Spotify now, play music, search songs, manage your playlist all through your AI [0:31] agent. And all of this is free. There were 213 people that worked on this update, 360 bugs, 1,096 improvements made. This is not a…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Julian Goldie SEO
NEW Nvidia Autonomous AI is WILD!! 🤯
Nvidia announced Nemo Clo, a new autonomous AI agent system that operates independently without continuous prompting. Powered by Nemotron 3 Ultra (a 550 billion parameter model), the system is five times faster and cheaper than previous versions, with OpenShell providing secure sandboxed execution.
Laguna XS 2.1: New FREE + Opensource Local AI!
Julian reviews Laguna XS 2.1, a new free open-source local AI coding model from Poolside that performs comparably to Qwen 3.6 and outperforms Claude Haiku on benchmarks. He demonstrates its practical capabilities by building landing pages and functional apps, highlighting its speed, offline functionality, and multiple deployment options through local setup, Claude Code, or OpenRouter's free API.
How to Run Hermes FREE Forever!
The video demonstrates how to run the Hermes AI agent for free using Gemma 4, a local open-source model from Google, with significant speed improvements through MLX optimization. The setup works on Apple Silicon Macs or via free APIs on Open Router, enabling autonomous agents to work offline and privately without subscription costs.
This NEW Chinese AI is INSANE! (FREE + Open Source!)
Long Cap 2.0 is a new open-source Chinese AI model from a food delivery app company that offers 1 million tokens of free context memory, beats GPT-4.5 on SWE bench pro benchmarks, and uses efficient parameter activation to reduce computational overhead while maintaining high performance.
Claude Code is now FREE: Here’s how…
Google's new Gemma 4 model running on Ollama is 90% faster on Apple Silicon, enabling free Claude Code usage locally without token costs. The setup requires three simple steps: downloading Ollama, Gemma 4, and installing into Claude Code, with alternatives available via OpenRouter API for non-Mac users.