Warp Sciences raises $215 million to build a chemistry lab and use its data to train AI and robots

Warp Sciences raises $215 million to build a chemistry lab and use its data to train AI and robots

On October 22, Kevin Weil, former chief product officer and vice president of science at OpenAI, announced the launch of Warp Sciences. The company is building a contract research organization in the US: a laboratory that carries out chemistry work for clients’ research programs and gradually automates that work.

Warp’s first service will be analog synthesis. Once a client has identified a promising molecule, chemists will make a series of closely related variants. The client can then compare their properties and select a candidate for further development.

The company’s announcement says clients will choose which molecules to pursue and retain control of their drug development programs. Warp will synthesize and test the compounds they need. Each project will follow a recurring cycle: make a variant, test it, and use the result to guide the next selection.

Experienced chemists will initially conduct the experiments. Warp then plans to automate more operations using AI and robotics. According to TechCrunch, the company plans to record chemists’ actions using cameras and sensor-equipped gloves to train robots to perform laboratory procedures. Warp also intends to use results from physical experiments to train AI models.

“Chemistry and biology advance when we make things, test them, and learn from the results. Building stronger AI for science requires real-world data at scale,” says Kevin Weil.

GPT-5 has already completed an automated cycle in Ginkgo’s robotic laboratory, working through six rounds and testing more than 36 thousand biochemical reactions. Warp is building a laboratory that provides chemistry services, with the aim of using results from client projects as training data for future automation.