OpenAI slashes 2030 compute budget to $600B, doubles revenue forecast to $280B

OpenAI’s decision to halve its 2030 compute budget reflects a strategic recalibration in response to spiraling hardware costs and a more aggressive monetization plan. The Manila Times reported that the company now projects total compute spend of roughly $600 billion through 2030, down from the $1.4 trillion target announced last year. The revision, echoed by the Indian Express and CryptoRank, signals that OpenAI is tempering its “scale‑first” ambition and reallocating capital toward revenue‑generating initiatives rather than pure compute expansion. Analysts familiar with the internal budgeting process, as cited by QUASA Connect, note that the shift coincides with a “bullish” outlook on product uptake, suggesting the firm believes its existing infrastructure can support the next wave of enterprise contracts without the previously envisioned exponential hardware outlay.

The revised compute outlook is paired with a dramatic upgrade to the company’s revenue forecast. According to a Reddit thread that references a The Information report, OpenAI has lifted its 2030 revenue projection to more than $280 billion, effectively doubling the figure it offered a few months ago. The same source adds that the company now expects an additional $111 billion in cash burn through 2030, a rise that mirrors the higher operating expenses tied to expanded sales, marketing, and data‑center capacity. The forecast upgrade underscores the confidence that OpenAI’s suite of products—ChatGPT, the GPT‑4o model unveiled in May, and its API services—are gaining traction among Fortune‑500 customers, a trend documented in recent ZDNet coverage of cross‑industry AI safety collaborations.

Revenue growth is being driven by a surge in enterprise adoption, a point highlighted in multiple outlets. Wired’s coverage of GPT‑4o notes that the model’s “snappy, flirty” interaction style and multimodal capabilities have broadened its appeal beyond consumer chat, prompting larger contracts for customized deployments. VentureBeat’s feature on OpenAI’s new “truth‑serum” training technique further illustrates the company’s push to differentiate its offerings on safety and reliability—attributes that corporate buyers increasingly demand. By coupling these product innovations with a more disciplined compute spend, OpenAI aims to improve margins that were squeezed last year when compute costs outpaced revenue, as QUASA Connect observed.

The financial recalibration also has implications for OpenAI’s competitive positioning. While rivals such as Anthropic, Google DeepMind, and emerging open‑source collectives continue to chase the same high‑performance hardware, OpenAI’s willingness to curb its compute ceiling may allow it to allocate more resources to partnership development and regulatory compliance—areas highlighted in the joint AI‑safety warning issued by researchers from OpenAI, Anthropic, Meta, and Google on ZDNet. In practice, a lower compute ceiling could reduce the firm’s exposure to supply‑chain bottlenecks that have plagued the semiconductor industry, thereby granting it greater flexibility to negotiate favorable terms with cloud providers and to invest in proprietary ASICs that lower per‑token costs.

Ultimately, the $600 billion compute target and $280 billion revenue outlook represent a balancing act between ambition and sustainability. By scaling back the raw compute budget while betting on higher‑margin enterprise sales and safety‑focused innovations, OpenAI is attempting to preserve its first‑mover advantage without overextending financially. If the company can deliver on the promised revenue growth, the revised plan could set a new benchmark for how AI‑first firms manage the trade‑off between massive infrastructure spend and profitability—a question that will dominate boardrooms and investor briefings through the remainder of the decade.