Nous Research Ships Hands-Free 'Hey Hermes' Wake Word to Its 175K-Star Agent
Hermes Agent adds always-on, fully local 'Hey Hermes' wake word detection across CLI, TUI, and desktop app
- Wake word shipped: Hermes Agent now supports hands-free "Hey Hermes" activation across CLI, TUI, and desktop app.
- Fully local: Detection runs entirely on-device; no audio leaves your machine until you speak a command.
- Three engines: openWakeWord (default, bundled model), sherpa (open-vocabulary, zero training), and Porcupine (custom .ppn files).
- Multi-profile routing: With the sherpa engine, saying "hey coder" or "hey trader" live-switches to the matching profile -- one listener, any profile.
- Off by default: Enable with
/wake onin CLI or the ear icon in the desktop app; persists toconfig.yamlautomatically. - Free and open-source: MIT licensed; available now at hermes-agent.nousresearch.com.
Hermes Agent just got a lot more ambient. Nous Research shipped wake word support, letting you trigger the agent hands-free by saying "Hey Hermes" -- no keyboard, no mouse, no context-switching. It works across the CLI, TUI, and desktop app, and the whole detection pipeline runs on-device.
What actually shipped
The wake word turns Hermes into a hands-free assistant: with one setting on, Hermes listens in the background for a spoken trigger phrase. Say it, and Hermes starts a fresh session, opens the microphone, captures your command via the normal voice pipeline, and answers -- exactly like "Hey Siri" or "Alexa".
Detection runs entirely on-device. The always-on listener only watches for the wake phrase; no audio leaves your machine until you actually speak a command to the agent. It is off by default, so nothing listens until you explicitly turn it on.
Three engines, all free
The feature ships with three detection backends you can swap between in config:
- openWakeWord (default) --
free, local ONNX models. Ships a bundled "hey hermes" model by default; also supports
hey_jarvis,alexa,hey_mycroft, and custom models. - sherpa -- open vocabulary: detects ANY typed phrase with zero training. A small English model auto-downloads on first use at roughly 13 MB.
- Porcupine --
the Picovoice engine; built-in keywords plus custom
.ppnfiles. Requires aPORCUPINE_ACCESS_KEYand has a free tier.
The open-vocabulary trick is the real story
Most wake word systems require you to train a model on your custom phrase -- a process that can take hours of compute. The sherpa engine sidesteps this entirely. Typed phrases need no training, and phrases such as "hey coder" can open a fresh session in the matching profile.