Prentis AI lab raises $100M at $1B valuation for computer-use agents
New startup aims to automate office workflows with cheaper, smaller AI models
Prentis, a new AI research lab launched in April, is in talks to raise $100 million at a $1 billion valuation. Co-founded by serial entrepreneur Ritankar Das (founder of Titan) along with LinkedIn co-founder Reid Hoffman and Zynga founder Mark Pincus, the lab focuses on building AI agents that can control computers to automate routine office workflows. Its Hive-32B model targets tasks like handling insurance claims and automating customs duty refund exceptions without human intervention.
Prentis claims its model outperforms OpenAI's GPT-5.4 and Anthropic's Claude Opus 4.6 on two benchmarks: WindowsAgentArena for end-to-end Windows task completion, and ScreenSpot-v2 for on-screen control accuracy. The startup says its edge comes from a smaller, cheaper model that reduces cost per task by roughly 10x compared to frontier APIs. It has already signed contracts worth up to $50 million with customers including a healthcare management organization and manufacturers. The company projects a $75 million annualized run rate by Q3 this year. Prentis faces competition from Anthropic, OpenAI, and Mira Murati's Thinking Machines Lab, but has hired over 25 researchers from top AI labs.
- Prentis is raising $100M at a $1B valuation, led by co-founders Ritankar Das, Reid Hoffman, and Mark Pincus.
- Its Hive-32B model claims to outperform GPT-5.4 and Claude Opus 4.6 on computer-use benchmarks while costing 10x less per task.
- The startup has already signed up to $50M in contracts with healthcare and manufacturing clients, targeting routine office workflow automation.
Why It Matters
If successful, Prentis could dramatically reduce enterprise costs by automating repetitive office tasks, challenging larger AI labs.