OpenAI Offers Up to $500,000 Base Salary for Robotics Roles

OpenAI is advertising a robotics role with a base salary of up to $500,000, before equity and bonuses, as the ChatGPT-maker scales its push into physical machines.

The San Francisco-based startup had 27 robotics jobs open on its careers page at the time of reporting, up from 11 openings in May, according to Business Insider. The listings span hardware, data collection, software and prototyping.

Publicly listed base pay across the openings ranges from $177,000 to $500,000. One robotics software engineer position in San Francisco pays $255,000 to $325,000 in base pay, plus undisclosed equity and performance-related bonuses. That role requires at least five years of professional software engineering experience developing systems in robotics.

The top-compensated opening is a machine learning engineering position focused on systems that handle robotics data at scale. The job involves building the underlying infrastructure to process, store and move training data across large networks of computers. It carries a listed base salary of up to $500,000.

OpenAI pairs the pay with generous benefits: a 401(k) retirement plan with employer match, paid parental leave of up to 24 weeks, and daily meals in the office.

What the postings reveal

Cornell mechanical and aerospace engineering professor Guy Hoffman told Business Insider that the postings indicate OpenAI is assembling a "custom robot design team." The company is recruiting engineers to develop robot hardware, including the motors and mechanical systems that power movement.

The openings also include a lab technician who would build, test and refine robotic systems, and an in-house lawyer assigned specifically to the robotics group.

One job asks a candidate to manage OpenAI's data collection facilities. The listing points to a structural difference between robotics and language models. Large language models learn from the massive troves of text and images available online. Robots need real-world examples of people, objects and machines interacting in physical environments — data that is harder to find, so robotics companies must frequently capture it themselves or buy it from specialized providers.

A long-term bet on machines

The language in the listings frames robotics as a major long-term bet. One posting describes the team's mission as advancing general-purpose robots and working toward human-like intelligence across multiple types of robotic machines. Another lays out a broader aspiration: a future in which "everyone" has a personal robot "doing anything they need."

OpenAI CEO Sam Altman has become increasingly direct about the company's plans to move from software into physical machines. In an interview on the podcast Sources earlier this month, Altman said OpenAI will "definitely do a humanoid" and "other form factors as well" — other robot designs suited to different jobs.

Altman pointed to data centers as a likely early use case, suggesting purpose-built robots could help operate and maintain the enormous computing infrastructure powering AI. Longer term, he described a consumer vision in which people own personal robots capable of handling a wide range of everyday tasks.

The hiring trajectory — more than doubling the robotics headcount openings in a matter of months, with data infrastructure commanding the highest pay — signals where OpenAI expects the next bottleneck in AI to sit: not in models, but in machines gathering experience in the physical world.