OpenAI and Synopsys ink a multi-year deal to build GPT-Synopsys, an AI model that designs chips
On Wednesday, Synopsys and OpenAI announced a multi-year strategic partnership to jointly develop GPT-Synopsys, a specialized AI model built to operate Synopsys' chip-design software the way expert engineers do. OpenAI will license Synopsys' EDA tools to build it, the deal includes revenue sharing and a joint go-to-market push — and it puts OpenAI directly in the business of designing the chips that train AI.
AI is coming for the software that designs AI hardware. In a joint announcement on Wednesday, Synopsys and OpenAI said they have signed a multi-year agreement to develop and deliver GPT-Synopsys, a specialized model optimized to use Synopsys' electronic design automation (EDA) tools to perform semiconductor design workflows. The companies will collaborate as preferred partners, OpenAI will license Synopsys' EDA tools for development of the model, and the agreement includes a revenue-sharing arrangement plus joint go-to-market collaboration to sell GPT-Synopsys to customers worldwide. Financial terms were not disclosed.
The pitch goes a step beyond today's agentic workflows. Right now, general-purpose models get bolted onto EDA tools with custom agent harnesses. The next leap, the companies say, is to make frontier models expert users of the tools themselves: learning to run them as veteran engineers do, interpreting their outputs, and iteratively optimizing designs. Engineers delegate design objectives — from power, performance and area (PPA) optimization to timing and verification closure — while agents run the tools, interpret results, implement changes, and iterate toward verified outcomes that engineers then review.
The quotes from both CEOs point at the same flywheel. Synopsys president and CEO Sassine Ghazi said the agreement "will expand access to Synopsys' advanced design capabilities and the underlying, ground-truth engineering tools required to bring increasingly complex chips to market." OpenAI president and co-founder Greg Brockman said OpenAI is using its most advanced technology to improve the systems that power AI — and that better chips will, in turn, let OpenAI build better AI for more people.
How it will ship#
GPT-Synopsys will run on OpenAI-hosted infrastructure, interoperate with customer agent harness systems, and integrate deeply with Synopsys.ai and the Synopsys Autopilot agentic AI platform. Early technology engagements are already underway with leading semiconductor customers, the companies said — so this is not a research memo, it is a product line in motion.
Data protection gets its own paragraph in the announcement, and for good reason: chip designs are crown-jewel IP. The joint offering will bundle compute, the model, and licenses with enterprise-grade security, governance, and access controls. Customer data will not be used to train the model, will be encrypted at rest and in transit, and can be managed through configurable retention, audit, and permission controls.
The context#
The deal lands two days after Synopsys introduced the Autopilot platform, its open and secure foundation for autonomous engineering, with a portfolio of domain-specific AgentEngineer solutions for verification, implementation, analog design, manufacturing, and simulation. Synopsys says more than 50 engagements are underway with AgentEngineer and the Autopilot Platform, with availability planned for the end of 2026. GPT-Synopsys slots in as the frontier-model brain on top of that plumbing.
Investors liked what they read: Synopsys shares rose 4.19% on the day to $435.50 on the announcement. And the strategic logic runs deeper than one stock pop — whoever controls the models that design chips gains leverage over the entire AI supply chain, from silicon to software. With GPT-Synopsys, OpenAI is now building on both sides of that stack.
