Hermes One

Hermes One is a community-maintained native desktop app for installing, configuring, and chatting with Hermes, the self-improving agent with tool use, multi-platform messaging, and a learning loop. Instead of managing the Hermes CLI by hand, you go through install, provider setup, and daily use in one GUI covering chat, sessions, profiles, memory, skills, tools, scheduling, and messaging gateways. It is not affiliated with the Hermes project and depends on it for agent behavior and tool execution.

On first launch the app asks whether to run Hermes locally or connect to a remote Hermes API server. In local mode it detects an existing installation or runs the official Hermes installer with dependency resolution, and in remote mode it validates a URL and API key and skips the install. It then asks for a model provider or local model endpoint and saves the configuration through Hermes's own config files. Chat runs over a streaming connection, and the app renders tool progress, Markdown, and token usage as the stream arrives.

Supported providers include OpenRouter, Anthropic, OpenAI, Google Gemini, xAI Grok, Nous Portal, Qwen, MiniMax, Hugging Face, and Groq, plus any OpenAI-compatible endpoint, with presets for LM Studio, Atomic Chat, Ollama, vLLM, and llama.cpp. API keys sit in a plaintext env file by default, and an opt-in command provider resolves them instead from a helper you configure, such as pass, secret-tool, Bitwarden, 1Password, or GnuPG, with a timeout and output cap on the helper.