Hark launches Handoff, a browser agent that automates online tasks

Hark, a startup that secured $700 million in Series A funding in May, has introduced its new agent, Handoff, which can use a browser to complete tasks efficiently. The company says Handoff can easily navigate websites that lack official APIs, including Target, Walmart, OpenTable and LinkedIn. The agent examines a site's structure and visual information to determine whether it needs to click buttons or type details.

The concept resembles many earlier browser-based agents: a user gives a command, and the agent carries out tasks such as ordering food or coffee, booking travel tickets, filing returns, buying essentials, reserving a restaurant table, or conducting research across multiple sources. In a video demo, CEO Brett Adcock showed the assistant following a command to build a bouquet with flowers specified by the user, while also handling vague phrases like "some of the florist's choice." However, the video only shows part of the process, so its actual effectiveness cannot be fully judged.

For this release, Hark says it uses a post-trained model and plans to move to pre-training later this year. The company believes this approach lets it improve its data pipeline, training infrastructure and techniques more quickly. Hark also claims that, unlike large language models that predict the next token, its model predicts the next action — which could be a click or a keyboard input at a specific location.

Several companies are working on computer-use agents, including Google, OpenAI and Anthropic, as well as VC-funded startups focused on browser task automation, such as Browser Use, Polar, Strawberry and Aside. Hark says Handoff is faster than the competition and costs significantly less than models like GPT 5.5 and Opus 4.8. The startup has opened a waitlist for its platform and expects to release it by the end of the summer.