Teknium's Manim Video Skill Goes Viral on Hermes Agent

Teknium posted the day's most-engaged Hermes tweet — a demo of Hermes Agent generating a full explainer video with the Manim Video skill and TTS tool, pulling 108 likes, 9 retweets, and 3,100+ impressions in under two hours.

In a follow-up reply, Teknium clarified that the Manim skill ships with every Hermes install — users just ask their agent to create a video on any topic, or use /manim-video <prompt> to force-load the skill. The reply collected another 10 likes and 688 impressions.

HYRE Gateway Ships 11-Model Routing

Hyre published a "More shipped" update confirming that the HYRE LLM Gateway — 11 models behind one OpenAI-compatible endpoint — now has a formal Hermes Agent integration. Users point Hermes at the Hyre endpoint and switch models from the gateway dashboard with no config rewrites.

Migration Stories Continue

@diego_defi reported leaving ChatGPT and Codex entirely in favor of Hermes with Nous Portal and OpenCode Go, citing cost savings from using more economical models through Hermes's model-agnostic runtime.

Notable Mentions

@SpitfingersX: "I've only heard good things about Hermes" (2♥). @ivanfioravanti: "I throw anything at my Hermes Agent powered by M3, no limits!" (1♥). @HermesAgentTips: "pretty interesting.. hermes is awesome" (1♥).

@D0m4nt45 noted they run a 45-tool production agent system on an 8GB GPU via Unsloth Dynamic 4-bit GGUF: "the hardware ceiling people imagine is way higher than the actual floor you need." @oceanavekid reported that Hermes "actually learns as it does tasks without me having to input anything" after switching from a Claude-based bot setup.

The PM window shows three patterns: the Manim skill demo from @Teknium dominated engagement (83% of all likes), infrastructure integrations with Hermes continue from providers like Hyre, and the steady drumbeat of users migrating from commercial chat interfaces to Hermes's model-agnostic runtime continues — now with specific cost savings being cited as the driver.