September 9, 2026, (Inside AI) — OpenAI is making a direct play for enterprise customers in chip design, life sciences, and financial services, with CFO Sarah Friar claiming its lower-cost models now beat Chinese open-source rivals on price when deployed through cloud providers.
Speaking at the Goldman Sachs Communacopia + Technology Conference in San Francisco on Monday, Friar said the company is experimenting with outcome-based pricing as enterprises demand measurable return on investment from AI spending.
The push targets sectors where specialized AI can deliver immediate value. Chip design is a notable test case. Friar said OpenAI used its own models to develop its Jalapeno chip, which reached tape-out within nine months. Tape-out is the stage when a chip design is finalized and sent to a factory for production.
OpenAI’s aggressive pricing strategy comes as competition intensifies from Chinese open-weight models and rivals such as Anthropic. Friar said OpenAI recently cut the price of its lower-cost Luna model by 80%, driving a roughly 10-fold increase in usage.
“If you’re deploying Luna and compare that to (Z.ai’s) GLM 5.3, for example, on a cloud layer, we are cheaper,” Friar said.
Open-source and open-weight models are widely seen as cheaper alternatives to frontier models from OpenAI and Anthropic. But Friar argues the total cost of deployment shifts the balance. Running an open-source model through a cloud provider involves infrastructure, maintenance, and optimization costs that can erase the upfront savings.
OpenAI’s Codex coding tool now has 25 million users, signaling strong developer adoption. Enterprise revenue increased 32% from June to July, compared with 20% growth in overall annualized revenue during the same period.
Enterprise and consumer businesses reached roughly an even split by mid-year, ahead of OpenAI’s target of reaching that balance by year-end. This shift reflects a broader industry trend where AI vendors are moving beyond generic chatbots to domain-specific solutions.
Outcome-Based Pricing Tests Enterprise Appetite
OpenAI’s experiment with pricing based on business outcomes rather than usage marks a significant departure from traditional software-as-a-service models. If successful, it could reshape how enterprises evaluate AI vendors.
The approach addresses a core enterprise complaint: AI spending has been hard to justify without clear return on investment. By tying price to outcomes, OpenAI is betting that customers will pay more when results are measurable.
Industry analysts note that outcome-based pricing has been tried before in enterprise software, with mixed results. The challenge lies in defining and measuring outcomes fairly for both parties. OpenAI has not disclosed specific metrics or pricing tiers for this model.
Competitors are watching closely. Anthropic has focused on safety and reliability for enterprise deployments. Chinese open-weight models like GLM 5.3 from Z.ai offer transparency and customization that some enterprises prefer.
The chip design use case is particularly strategic. Semiconductor companies face immense pressure to accelerate design cycles. If OpenAI can consistently reduce tape-out timelines, it could become an essential tool in the chip industry.
OpenAI’s internal use of its models for the Jalapeno chip serves as a proof point. But independent validation from third-party chip designers will be necessary before widespread adoption.
Cloud Economics Shift the Competitive Landscape
Friar’s claim that Luna undercuts Chinese open-source models on cloud deployment costs challenges a core assumption in the AI market. Open-weight models have been marketed as the budget-friendly option.
The comparison hinges on total cost of ownership. Cloud providers charge for compute, storage, and data transfer. Optimizing an open-source model for production requires engineering talent and ongoing maintenance. These hidden costs can exceed the license savings.
OpenAI’s managed API approach bundles these costs into a predictable per-token or per-outcome price. For enterprises without deep AI engineering teams, this simplicity has value.
However, open-source advocates argue that self-hosting eliminates vendor lock-in and data privacy concerns. Some regulated industries, such as healthcare and finance, may prefer to keep models on their own infrastructure.
The enterprise AI market is still in flux. OpenAI’s pricing moves and vertical focus suggest a maturing strategy aimed at capturing high-value, repeatable revenue streams. The coming quarters will reveal whether outcome-based pricing gains traction beyond early adopters.
