Nous Portal is 📝Nous Research's unified subscription gateway, a single account that supplies 📝Hermes Agent with model inference and hosted tools, and the docs' recommended way to run the agent.
The Hermes Agent docs describe a catalog of 300+ models billed against one Nous subscription, and the Portal's own page says it spans hundreds of models from frontier labs. The docs' snapshot includes Anthropic Claude, OpenAI and Google Gemini models. The Portal routes each model to the backend best suited for it, some through 📝OpenRouter and others through proprietary or secondary providers, and that routing can change over time.
The same subscription unlocks the Tool Gateway, which routes Hermes Agent's tool calls through Nous-managed infrastructure: web search and extraction, image generation, text-to-speech, cloud browser automation and an optional cloud terminal sandbox. The gateway is opt-in per tool.
Setup is one command, hermes setup --portal, which runs the OAuth login, writes Nous as the inference provider with base URL https://inference-api.nousresearch.com/v1 into config.yaml, and turns on the Tool Gateway. The refresh token at ~/.hermes/auth.json is the only credential on disk; Hermes mints a short-lived JWT from it on each inference call instead of reusing a long-lived API key.
Nous's own Hermes 4 models, covered in 📝Nous Model Family, are available through the Portal, but the docs do not recommend them inside Hermes Agent: they are tuned for chat and reasoning, not the rapid tool-calling loop the agent relies on.