A new AI lab you probably haven’t heard of just stepped into the spotlight with a billion-dollar price tag. Prentis, a research startup focused on computer-use models, is in talks to raise $100 million at a $1 billion valuation, according to TechCrunch AI, which cited two people familiar with the discussions. The company was co-founded by serial entrepreneur Ritankar Das alongside two of tech’s most recognizable names: LinkedIn co-founder Reid Hoffman and Zynga founder Mark Pincus.
What stands out here is how fast this happened. Prentis only launched in April. A billion-dollar valuation and a nine-figure raise inside a few months tells you how hot the computer-use category has become.
What Prentis actually builds
Prentis trains AI models to watch how office workers move through routine tasks: navigating documents, clicking through systems, chasing paperwork. The goal is AI agents that can control a computer and do that work on their own. Think handling insurance claims or automating customs duty refund exceptions without a human digging for forms.
The company isn’t just pitching a vision. According to TechCrunch AI, Prentis has already signed contracts worth up to $50 million with several customers, including a healthcare management organization and a set of manufacturers. Investor materials project a $75 million annualized run rate by the third quarter of this year.
One important caveat: Prentis’s own pitch deck notes those numbers reflect estimated annualized value based on a fee equal to 20% of savings realized, not recognized revenue. They’re “performance-dependent and subject to final execution.” So treat the headline figure as a target, not money in the bank.
The technical claim
Prentis says its Hive-32B model beats OpenAI’s GPT-5.4 and Anthropic’s Claude Opus 4.6 on two computer-use benchmarks:
- WindowsAgentArena, which measures end-to-end task completion on real Windows apps.
- ScreenSpot-v2, which tests whether a model can find the right on-screen control.
The interesting part is the strategy behind it. Prentis argues its edge comes from running a smaller, cheaper model rather than a giant frontier one. The company claims roughly 10 times lower cost per task than frontier APIs, which it says makes the model economical to deploy across everyday work. TechCrunch AI notes it hasn’t independently verified those benchmark results, so keep a healthy dose of skepticism.
This is significant because cost per task is exactly where office automation lives or dies. A model that’s slightly less capable but ten times cheaper can win contracts a frontier model can’t touch.
Why this matters
Prentis is betting that automating routine office work will soon overtake coding as AI’s biggest use case. That’s a bold call, and it’s a crowded lane. Anthropic, OpenAI, and Mira Murati’s Thinking Machines Lab are all chasing computer-use agents. Anthropic has been buying talent outright, snapping up Seattle startup Vercept earlier this year and shutting down its product.
For practitioners, the takeaway is clear: the agent wars are shifting from writing code to running the back office. If Prentis is right, the next battleground is the boring, repetitive work that fills most companies’ day.
The people behind it
Ritankar Das, 31, is the CEO. He was UC Berkeley’s youngest University Medalist in over a century, graduating at 18, and later founded Titan, a holding company he’s described as a throwback to the Berkshire Hathaway model, funded by its own exits rather than outside investors. Titan’s other bets include Tala Health and autism-care startup Forta Health.
Hoffman and Pincus are more side players here. Hoffman recently said he’s going “founder mode” on a separate drug-discovery startup, and Pincus runs investment firm Reinvent Capital. Their names carry weight, but Das is clearly driving.
Prentis has already hired more than 25 people, including researchers from OpenAI, Google DeepMind, Meta, Tencent, and Alibaba. Prentis didn’t respond to TechCrunch’s request for comment.
Watch this space. If the raise closes and those contracts convert, Prentis could become one of the defining names in enterprise AI agents. Full details are available at the original TechCrunch AI report.