Yutori

What Yutori does

Yutori is building AI agents that can reliably perform everyday digital tasks on the web, aiming to create an AI chief-of-staff for everyone. These agents can handle simple tasks like ordering groceries or tracking campground reservations, as well as more complex scenarios such as planning and coordinating group trips. Yutori is reimagining web interaction by building an agent-first stack, which includes training their own models and developing generative product interfaces. They offer various APIs (Navigator, Browsing, Research, Scouting) to enable web automation, research, and continuous monitoring.

Products

• NavigatorYutori's most capable computer-use model, designed to interact with webpages in browser environments. It processes user instructions, current screenshots, and action history to predict the next action (click, type, scroll, etc.). The latest model is Navigator n1.5.
• Browsing APIEnables automation of browser-based workflows by running Navigator models on Yutori's cloud browser. Users describe tasks in natural language, and an AI agent operates the browser to complete them.
• Research APIEnables one-time, comprehensive research on the web. Users specify research topics in natural language, and the API uses over 100 MCP tools for execution.
• Scouting APIEnables continuous monitoring of the web at a configurable schedule for tracking changes relevant to a query. A 'Scout' spins up sub-agents to monitor specific URLs or the entire web and alerts with structured data upon relevant updates.
• Yutori Python SDKA software development kit for calling all Yutori APIs.
• MCP serversServers available for calling all Yutori APIs.
• Yutori LocalA Mac desktop app that provides Yutori agents with access to a browser on the user's computer, enabling login-required workflows and local browsing tasks.
• DelegateAn AI agent for founders that handles busywork such as investor updates, feedback synthesis, partner follow-ups, intro requests, and competitive intelligence.

Key capabilities

• Autonomous web task execution
• Agent-first stack
• Proprietary web navigation models
• Generative product interfaces
• Cloud browser for automation
• Multi-agent research capabilities
• Continuous web monitoring and alerts
• Python SDK
• Mac desktop app for local browsing tasks
• API usage monitoring and task history management platform

Use cases

• Ordering groceries
• Tracking campground reservations
• Planning and coordinating group trips
• Filling forms on websites
• Checking webpages for changes and retrieving updated information
• Checking multiple webpages and downloading summaries in structured formats
• Logging into websites and entering or updating information
• Researching latest developments in quantum computing
• Summarizing recent AI model launches and their capabilities
• Researching competitor pricing for cloud GPU instances
• Tracking startup seed funding announcements
• Monitoring H100 pricing drops
• Tracking new AI model launches
• Generating monthly investor updates
• Synthesizing weekly user feedback
• Planning beta launch outreach
• Tracking and facilitating introductions
• Monitoring competitor activities (shipping, hiring, press, funding)
• Tracking team deliverables and follow-through

AI approach

Yutori builds AI agents that autonomously browse and act on the web, aiming to delegate everyday digital tasks. They are building an agent-first stack, including training their own web navigation models and inventing new products enabled by background web agents. Their Navigator models interact with webpages in browser environments, processing user instructions, screenshots, and action history to predict next actions.

Tech named: Navigator n1.5, claude-sonnet-4-5 (CU), MCP tools, Yutori Python SDK, MCP servers

What it says sets it apart

• Building the entire stack to be agent-first
• Training their own web navigation models
• Inventing new products enabled by background web agents
• Accuracy-cost-latency Pareto dominance over frontier models (Navigator n1.5)
• Uses 100+ MCP tools for comprehensive web-based research (Research API)
• Ability to monitor specific URLs or the entire web for changes (Scouting API)
• Dedicated capacity, SLAs, and on-prem deployment available for high-volume/enterprise customers

Funding rounds we track

From the AI funding tracker — rounds as reported by the linked publications.