Your AI. Your choice of where it runs.

Choose between the browser, your local model server and cloud AI from one workspace.

Your Browser

Run a model on a compatible computer, phone or tablet directly from the website. No app installation or personal signup is needed. After downloading a model, Your Browser processes prompts on that device. Graphics support and available memory determine which models fit.

Your Machine

Connect an existing local model server such as Ollama or a compatible local API. Noema must be able to reach that server from the browser you are using. A server on a desktop is not automatically reachable from a phone.

Noema Router

Use private cloud AI when you need more capacity, including on mobile. Auto uses zero-retention routes; ZKR masks detected sensitive text when enabled. No personal signup or separate provider keys are needed. Direct model choices show their own prices and retention labels. Credit activity is wallet-linked.

Keep costs and context visible

Local model execution uses your own hardware. Cloud requests consume credits. Optional web search and external tools have their own network activity, even if the model itself runs locally.

Open NoemaExplore the docs