Hermes Agent with Nbility
Hermes Agent’s model wizard supports custom OpenAI-compatible endpoints. A dedicated provider keeps the gateway separate from built-in provider logins.
Use the setup wizard
- Install Hermes Agent and run
hermes modelin your terminal. - Select Custom endpoint and enter
https://api.nbility.ai/v1. - Supply a dedicated API key through the prompt and select Chat Completions compatibility.
- Select or manually enter an exact ID from Models. Set the context length to the real model limit.
- Save, launch Hermes, and send a short message.
The terminal command hermes model runs full setup. The in-session /model command switches among configured models.
Review auxiliary models
Check Usage logs for the main conversation. Compression, vision, titles, and other auxiliary tasks can have separate model settings; review their provider and model if you also want them routed through Nbility.
A working model does not configure independent search, browser, or voice services. Follow each tool’s documentation and use its own credentials.
Troubleshooting
If discovery fails, check token access with the Models API and enter the model manually. For failed tools, verify model tool support and review context limits after changing models.
References
Adapted and rewritten for Nbility from aicentos’s Hermes Agent guide, licensed under CC BY 4.0. Its account, plan, and proprietary service instructions do not apply here. See the client’s official documentation.