OpenAI Touts Chip Design With Lower-Cost Strategy

OpenAI is expanding its AI push into chip design, positioning its models as tools for semiconductor development while arguing that its technology can compete on price with open-source and open-weight alternatives.

The company has already tested the approach on its own hardware. OpenAI used its AI models in the development of Jalapeño, an inference processor jointly developed with Broadcom. The chip reached tape-out in less than nine months, according to OpenAI CFO Sarah Friar. Tape-out is the stage when a chip’s design is finalized and sent to a fabrication facility for manufacturing.

Jalapeño is designed to run OpenAI models at lower cost. Early samples indicate the processor could reduce costs by roughly 50% compared with typical AI GPUs. OpenAI and Broadcom are targeting initial deployment by the end of 2026.

For OpenAI, its own chip project provides a working example as it seeks enterprise customers for AI-assisted semiconductor design. Chip development is a complex process with lengthy design cycles, making it a potentially valuable market for AI tools that can boost engineering work.

OpenAI is making cost a key part of its competitive strategy. Open-source and open-weight AI models have attracted enterprise interest because they can offer an alternative to proprietary frontier models. Yet businesses still need to pay the computing and cloud costs required to operate those models.

Friar argues that this changes the economics of the comparison. OpenAI recently cut the price of its lower-cost Luna model by 80%, a move that was followed by a roughly tenfold increase in usage.

Industry-Specific Workflows

The combination of chip design capabilities and lower model prices could give OpenAI another route into enterprise accounts as competition grows from Chinese open-weight models and other commercial AI providers. Rather than competing solely on model performance, OpenAI is attempting to demonstrate that its technology can handle specialized technical work at a competitive operating cost.

Chip design is part of OpenAI’s larger effort to develop AI around industry-specific workflows. The company is also targeting life sciences and financial services, where specialized business processes could provide opportunities for AI systems tailored to specific tasks.

That expansion comes as OpenAI reports strong growth from business customers. Friar said enterprise revenue rose 32% between June and July, compared with 20% growth in overall annualized revenue during the same period. By the middle of the year, the enterprise and consumer sides of OpenAI’s business were approximately equal in size, a milestone the company had expected to reach by year-end.

OpenAI’s Codex coding tool has become a key contributor to its business adoption. The tool reportedly now has 25 million users.

The company is also experimenting with pricing based on business outcomes rather than simply the amount of AI a customer consumes. That approach is a clear appeal to enterprise buyers that want a clearer connection between their AI bills and measurable business results.