OpenAI Projects Cumulative Losses of $278 Billion from 2026 to 2030
nashnova research
An internal presentation obtained by the Financial Times shows OpenAI expects $278 billion in negative free cash flow over the next five years — even as revenue grows tenfold, compute spending outpaces it, and the company seeks fresh funding at a valuation above $1.2 trillion.
Five-year revenue hits $840 billion — so where does it all go?
OpenAI projects $36 billion in revenue this year, rising to $350 billion by 2030 — roughly $840 billion over five years.
Compute and infrastructure alone will cost $856 billion over the same period — more than total revenue.
This means → ten-times revenue growth still cannot cover the single largest expense line. The faster OpenAI scales, the faster it burns cash.
How long does the $122 billion last?
OpenAI closed a $122 billion funding round in March. Internal projections show this capital runs out by 2028.
This means → the company will need to raise again within roughly two years. Fundraising is not a one-off event — it is a recurring cycle.
Investors have already approached OpenAI to discuss a new round at a $1.2 trillion valuation; the company is pushing for an even higher price.
Is the loss forecast narrowing or widening?
An earlier projection in May put five-year negative cash flow at $305 billion; the latest figure is $278 billion — a $27 billion improvement.
After a new model launch in July, annualized revenue grew roughly 20% quarter-on-quarter — the top line is accelerating.
In plain terms = the gap is still enormous, but the direction is improving. The question is whether improvement can outrun spending.
Why was the IPO postponed?
OpenAI had planned to go public this autumn and confidentially filed with the SEC in June, but later delayed the process.
The official reason: rising public concern over the risks of rapid AI development. Some investors say the market is uneasy about how to price a company still posting massive losses.
This reflects a broader reality — no matter how high the private valuation, public markets still want to see a path to profitability.
How does competition push burn rates higher?
OpenAI faces a two-front battle: competing with Anthropic for enterprise clients while fending off low-cost open-weight models — models whose parameters are published for anyone to use freely — from China.
This means → it cannot raise prices (clients would leave) yet must keep increasing compute spend (or fall behind). Margins are squeezed from both sides.
Its compute contracts are deeply tied to Nvidia, Oracle, and SoftBank-backed data centers — companies whose own revenue depends in significant part on OpenAI honoring those contracts.
市场有风险，内容仅供研究参考，不构成投资建议。