Simular is building autonomous computer-use agents that go beyond the browser to operate an entire desktop. Its agents control a user's Mac or Windows PC directly, navigating native applications, clicking, typing, and completing multi-step tasks the way a person would, which lets them tackle work that lives outside the web in the everyday software people actually use.
At the core of Simular's approach is Agent S, an open agentic framework for computer use. On OSWorld, a demanding benchmark for desktop agents, Agent S reached a 69.9% success rate, approaching the roughly 72% achieved by humans, a striking result that signals how close computer-use agents are getting to practical reliability on real operating systems. By open-sourcing the framework, Simular has positioned itself at the research frontier of desktop automation while building a product on top of it.
Computer use is widely seen as one of the most consequential frontiers in agentic AI because it unlocks automation of tasks that resist API-based integration, legacy apps, desktop tools, and workflows that span many programs. Simular's bet is that controlling the whole computer, rather than just a sandboxed browser, is the path to agents that can genuinely do a person's digital work.
Simular raised a $21.5 million Series A led by Felicis in December 2025, bringing total funding to $27 million. The round included NVentures (NVIDIA's venture arm), Basis Set Ventures, Flying Fish Partners, and South Park Commons, with angel Lenny Rachitsky and earlier backers Samsung NEXT and Xoogler Ventures. The founding team includes AI experts from Google DeepMind, Baidu, and UCSB, with a Best Paper Award at the ICLR 2025 Agentic AI workshop.
For users and enterprises seeking agents that can operate the full breadth of desktop software, Simular offers a research-grade, autonomous computer-use platform aimed squarely at the hardest part of automation.