Anthropic Acquires Vercept to Enhance Claude's Computer Use Abilities

Anthropic has acquired Vercept, a startup specializing in AI perception and interaction, to bolster Claude's computer use capabilities as the model approaches human-level performance on complex software tasks.

anthropic Feb 25, 2026

Users are relying on Claude for increasingly sophisticated work-writing and executing code across full repositories, synthesizing research from numerous sources, and orchestrating workflows that span multiple tools and teams. Computer use allows Claude to perform all of this within live applications, mimicking the way a human would operate at a keyboard. This means Claude can handle multi-step tasks in real software environments and tackle problems that code alone cannot solve. Anthropic has announced the acquisition of Vercept to further advance these capabilities.

Vercept was founded on a clear premise: making AI truly effective at completing complex tasks demands solving difficult perception and interaction challenges. The Vercept team-including co-founders Kiana Ehsani, Luca Weihs, and Ross Girshick-has spent years focused on how AI systems can observe and interact with the same software people use daily. That expertise aligns directly with some of the most challenging problems Anthropic is tackling. Vercept will be winding down its external product over the coming weeks as the team joins Anthropic to push the boundaries of computer use.

This acquisition follows the recent release of Claude Sonnet 4.6, which demonstrated a significant leap in computer use performance: on OSWorld, a widely-used benchmark for AI computer use, Anthropic's Sonnet models improved from under 15% in late 2024-when computer use was first introduced-to 72.5% today. Sonnet 4.6 is now approaching human-level performance on tasks such as navigating complex spreadsheets and filling out web forms across multiple browser tabs.

Vercept is the latest team brought into Anthropic, following the acquisition of Bun. Anthropic seeks out teams whose technical ambitions align with its own, whose work advances its capabilities, and whose approach to building AI is rooted in the same principles of safety and rigor that guide everything at the company.

Anthropic is also hiring individuals. Those interested in joining the engineering team can visit Anthropic's careers page.