Hermes · API Server
OpenAI-compatible endpoint. Point any client at it.
The API server exposes Hermes as an OpenAI-compatible HTTP endpoint. Any OpenAI-format frontend, Open WebUI, LobeChat, LibreChat, the OpenAI SDK, curl, can use Hermes as its backend, with the full toolset behind it. Enable with API_SERVER_ENABLED=true + API_SERVER_KEY in .env, then hermes gateway; it listens on http://127.0.0.1:8642. Point any client at /v1 with the key as a bearer token (OpenAI SDK: set base_url). Two endpoints: /v1/chat/completions (stateless) and /v1/responses (server-side state via previous_response_id or a named conversation). A frontend system prompt layers on top of the core prompt. Multi-user via profiles (own port + key, profile name as model). Security: full toolset incl terminal, key required off-loopback, CORS off by default.
Sources · What this video distills
1 docs page · every command below traces to one of themCommands shown · Copy and paste
each shows the source doc it came fromAPI_SERVER_ENABLED=true · API_SERVER_KEY=change-mehermes gateway # → listening on http://127.0.0.1:8642curl http://localhost:8642/v1/chat/completions -H "Authorization: Bearer change-me" -d '{"model":"hermes-agent","messages":[{"role":"user","content":"Hello!"}]}'OpenAI(base_url="http://localhost:8642/v1", api_key="change-me")