Hermes Workspace

Connecting to your backend...

Welcome! Let's connect your backend

Hermes Workspace works with any OpenAI-compatible backend. Hermes gateway APIs unlock enhanced features automatically when they are available.

1Use any OpenAI-compatible backend

Set HERMES_API_URL to your backend base URL

Portable chat works with any backend that exposes /v1/chat/completions (Ollama, LiteLLM, vLLM, etc.)

2Optional: run a Hermes gateway locally

git clone https://github.com/outsourc-e/hermes-agent.git

Hermes gateway APIs unlock sessions, skills, memory, and other workspace extras automatically

3Install the gateway

cd hermes-agent && python3 -m venv .venv && source .venv/bin/activate && pip3 install -e .

4Enable the HTTP API server

echo "API_SERVER_ENABLED=true" >> ~/.hermes/.env

The gateway HTTP API is opt-in. Without this, the gateway serves messaging platforms but does not expose port 8642 for the workspace.

5Start the gateway

cd hermes-agent && source .venv/bin/activate && hermes --gateway

Or use Auto-Start below if hermes-agent is already installed locally

Point HERMES_API_URL at any OpenAI-compatible backend:

HERMES_API_URL=http://your-server:8642 pnpm dev

This page auto-refreshes when a compatible backend is detected