Sam Altman, CEO of OpenAI, has publicly reaffirmed the company’s close partnership with AI chipmaker Cerebras Systems, addressing recent market speculation that suggested a cooling relationship. The statement, shared on the platform X, aims to clarify the ongoing collaboration despite reports of OpenAI’s newest model, GPT-6.1 Sol, initially deploying on Nvidia hardware.
OpenAI and Cerebras Systems announced a significant multi-year partnership on January 14, 2026. Under this agreement, OpenAI committed to purchasing up to 750 megawatts of Cerebras AI inference capacity over the following three years. The initial value of this deal was reported to be over $10 billion, with potential extensions that could push the total value beyond $20 billion. Cerebras pitches its hardware specifically for ultra-low-latency inference, designed to make AI responses faster. This capacity is intended to be integrated into OpenAI’s inference stack, aiming to enhance the responsiveness of applications like ChatGPT.
Speculation Fueled by Deployment Reports
The reaffirmation comes after market rumors surfaced in early October 2026, suggesting that OpenAI’s GPT-6.1 Sol model was initially deployed on Nvidia GPUs rather than Cerebras hardware. According to those reports, this development, coupled with Cerebras’s focus on ultra-low-latency inference, fueled speculation among investors about a potential shift in OpenAI’s strategy and a cooling of the partnership. Cerebras shares had reportedly come under pressure following these rumors, highlighting a structural vulnerability for the company due to customer concentration, as OpenAI represents a significant portion of its business.
Altman’s Dual Role and Investment
Altman’s personal investment in Cerebras, dating back to at least February 2017, adds another layer to the situation. As of May 2026, he held approximately 89,373 shares in the company, which went public in May 2026. His dual role as CEO of OpenAI, a major customer, and a longtime shareholder in Cerebras makes his public reassurance notable and may continue to prompt discussions about governance.
Complementary Hardware Strategy
Both OpenAI and Cerebras have framed the partnership as complementary. Cerebras’s dedicated, low-latency capacity is seen as sitting alongside OpenAI’s existing general-purpose GPU fleet from suppliers like Nvidia and AMD, rather than as a replacement. The agreement is multi-year, with the capacity rollout planned in phases through 2028. Analysts believe that the market may have overstated the risk associated with the initial deployment choice for a single model, emphasizing that a one-time deployment decision does not alter a three-year purchase commitment.
Uncertainties Remain
While Altman’s statement on X provides clarity, the exact timing of his post remains unspecified. The potential for the total deal value to exceed $20 billion is also subject to extensions, and the extent to which governance questions will be raised regarding Altman’s dual role remains to be seen. Despite these nuances, the core message from OpenAI’s CEO is that the partnership with Cerebras remains a key component of its infrastructure strategy.
Why This Matters
The materials describe a narrow update: Sam Altman publicly stated that Cerebras is a close partner of OpenAI, following market rumors suggesting a cooling relationship. The exact timing of Sam Altman’s reaffirmation on X.
Broader Context
Source materials place the factual news in this context: The reaffirmation, shared on X, follows a stretch of speculation about whether the relationship between OpenAI and the AI chipmaker was cooling.
