Polar secures $5.7 million to automate knowledge work in the browser
In 2026, AI-browser startup Polar raised a $5.7 million seed round led by Madrona. CEO Kevin Jiang previously worked on Comet at Perplexity. Polar is free with limited daily AI credits, while subscriptions with higher limits start at $20 per month.
Why Polar is avoiding the mass market
The first generation of AI browsers tried to replace default search and help consumers complete occasional tasks such as booking travel. Jiang argues that these activities do not recur often enough to give a mass audience a durable reason to switch browsers.
Polar instead targets sales, recruitment, marketing, research and business operations, where repetitive browser work is constant. Its strategy reflects the browser market's shift from search to AI agents and shifts the competition from answering questions to executing reliable workflows.
“If you think about end-user consumers, they don’t book a flight or make a reservation every day or every week. There wasn’t a strong pull for mass consumers to go to an AI browser and find value in it,” Jiang said.
What the browser can automate
Users can schedule workflows, save prompts and assign tasks to agents based on their open tabs. Polar says the product is accessible to nontechnical staff and can operate across the web services they already use.
Most current users keep another browser as their daily driver and deploy Polar primarily for automation. The startup wants to become an automation platform and considers the browser its best interface today, although it may also explore computer-use technology.
Reliability remains the decisive test
Madrona partner Sabrina Albert described automation for knowledge workers as a large market. She said the difficult part is delivering dependable agentic work across the open, logged-in web; a browser can observe interactions and access services where users already have active sessions.
The round also included former GitHub CEO Thomas Dohmke, Phoebe Gates, Erik Bernhardsson, Benjamin Spector, Rob Wachen and Robert Yang.
For businesses, the practical opportunity is not an immediate company-wide browser replacement. A safer approach is to select one bounded, repetitive workflow, measure completion accuracy and review costs, and expand only when the agent produces consistent savings without weakening access controls.