Hermes HQ Subscribe
Setup

Private AI models: run open models on your computer with no API key

What it does

Hermes Agent v0.22.0 automatically installs llama.cpp and detects your hardware (CPU, GPU, RAM) to select the best open model that fits. Models run entirely on your computer with no account, no API key, and nothing leaving your device.

Why it is worth it

Get complete privacy for your AI work and unlimited inference without API costs or external dependencies. Automatic hardware detection and model selection make local AI accessible without technical friction.

Related use cases

See all
Screenshot: Agent stuck on your desktop: text it from your phone, set up by QR code
SetupOct 12

Agent stuck on your desktop: text it from your phone, set up by QR code

The dashboard creates the Telegram bot for you and adds your account to the allow list. Scan, tap Connect, and your agent is in your pocket.

View use case
SetupOct 8

API bills for a private agent: it runs free on a 16GB Mac Mini

Gemma 4 12B through LM Studio, Honcho for memory, Telegram for remote access, and one Mac setting so it never sleeps. Daily AI news lands in Telegram.

View use case
SetupOct 11

A backup you have never tested: prove it restores without touching live files

A read-only check rebuilds your newest backup in a temp folder and reports names and counts. Your live setup and keys are never opened.

View use case