
Yutori is a web-agent platform for developers and people tracking online information. Scouts monitor the web, Navigator powers computer-use agents, and the Research API investigates questions across websites.
Yutori separates its computer-use model from managed browser execution and multi-agent research. Choose Navigator when building your own agent, Browsing API to complete a website workflow, or Research API to gather and synthesize findings across sites.
Browsing API takes a task and starting URL, then uses Navigator n1.5 to click, type, scroll, and extract data. It handles JavaScript-heavy and authenticated pages. Tasks run asynchronously, with a live view and optional structured output.
Research API returns cited findings and supports JSON output through an output schema. Scouts offer a web interface for monitoring topics; the Scouting API lets developers set a recurring schedule. Navigator n2 also supports broader desktop, terminal, and code workflows.
Yutori publishes computer-use benchmark results, which measure task performance rather than customer satisfaction. For a practical evaluation, check whether browser actions finish correctly and whether research citations support the conclusions. Test representative workflows before relying on unattended runs.
These rates apply to the Navigator model API. Check Scouts and hosted API billing separately before budgeting a deployment.
+2 more
+2 more