OpenAI and Synopsys have announced a partnership to build GPT-Synopsys, a specialised model intended to operate the electronic design automation (EDA) tools that engineers use to create chips.

The two companies say GPT-Synopsys combines OpenAI's frontier models with Synopsys' EDA technology and domain expertise, so the model can reason about design intent and act on it through the tools — turning an engineering objective into tool-driven work. Under the arrangement, OpenAI licences Synopsys' EDA tools to develop the specialised model; the companies will share revenue and collaborate on it. Early customer tests are already under way, and engineers are expected to review and approve the model's output.

The idea is not to replace chip designers but to compress the long, repetitive middle of the workflow — the scripting, iteration and option exploration that eats weeks of an engineer's time. Synopsys has been building in this direction for years, and separately announced an agentic AI stack with AMD and Microsoft, plus an "Autopilot" platform for autonomous chip development targeted at general availability by the end of 2026.

If it works, the significance is bigger than one product. Chip design is one of the most tool-locked, licence-gated workflows in engineering, and an AI that can drive those tools is a template for every other domain where value sits behind a specialised, expensive application. The near-term questions are the usual ones for agentic work in a high-stakes, error-intolerant field: how much of the flow can be trusted to a model, how the outputs get verified, and who is accountable when a bad suggestion reaches a tape-out. That is why the announcement comes with the qualifier that engineers review and approve.