How-To Series · Episode 50 / 59 · Module 8: Provider Plumbing

Hermes · The Provider Landscape

Three ways to power Hermes: a subscription gateway, your own keys, or a local model. Pick by wallet and work.

After this videoYou can now pick a provider strategy that fits your wallet and your work.

Every model runs through one of three buckets. A subscription gateway (Nous Portal, the recommended default): hermes setup --portal gives you 300+ models under one login, no per-provider keys. Your own provider keys: drop an OPENROUTER_API_KEY (or Anthropic, OpenAI, etc.) in .env and pay each provider directly. A local model: point Hermes at Ollama, vLLM, or llama.cpp via a custom endpoint for $0/token and full privacy. Set your default with hermes model (interactive provider + model picker, writes config); hot-swap a running chat with /model.

About these resources. Every command comes from the AI Providers doc; the Nous Portal, Provider Routing, and Fallback docs are cited for further reading.

Sources · What this video distills

1 docs page · every command below traces to one of them
Primary · the three buckets, Nous Portal, direct keys, custom/self-hosted endpoints, hermes model vs /model
AI Providers
Read ↗

Commands shown · Copy and paste

each shows the source doc it came from
Subscription (Portal)from source ↗
hermes setup --portal · hermes model
Your own keyfrom source ↗
echo "OPENROUTER_API_KEY=sk-or-..." >> ~/.hermes/.env
Local modelfrom source ↗
ollama serve → hermes model → Custom endpoint → http://localhost:11434/v1
Hot-swap mid-chatfrom source ↗
/model openrouter:anthropic/claude-sonnet-4

Going deeper · Related Hermes docs

further reading · not sources of facts shown above
Nous Portal
what's in the subscription
Provider Routing
pick OpenRouter sub-providers by cost/speed
Fallback Providers
survive provider outages

Next in the series · Episodes that build on this

E51
Run Local LLMs on Mac
E52
Provider Routing
E53
Fallback Providers