OpenAI spent $1.22 for every dollar it earned in 2024, even after removing stock-based compensation.
The company reported $3.7 billion in revenue against $4.5 billion in operating costs. That left an $800 million loss for the year.
Excluding stock-based pay, the gap narrows but still shows a 22 percent cost overrun. The figures come from internal financial documents reviewed by multiple outlets.
Revenue growth masks deep cash burn
OpenAI’s revenue doubled year-over-year, driven by ChatGPT subscriptions and API sales. But costs grew even faster.
The biggest expense is computing power. OpenAI spends heavily on Nvidia GPUs and cloud infrastructure from Microsoft.
Training and running large models consumes billions of dollars annually. The company projects it will need up to $10 billion more in funding this year.
Stock-based compensation hides real losses
When including stock-based compensation, OpenAI’s losses deepen significantly. Employees received equity grants worth roughly $1.5 billion in 2024.
Adding that back pushes total expenses above $6 billion. That means the company burned roughly $2.3 billion on a GAAP basis.
“Without aggressive stock grants, OpenAI’s cash flow would look even worse,” one analyst told reporters.
The company uses equity to attract top AI talent without spending cash. But that dilutes existing shareholders and masks operating reality.
Microsoft partnership provides lifeline
Microsoft has invested over $13 billion into OpenAI since 2019. In return, Microsoft gets exclusive access to OpenAI’s models for its cloud and products.
That partnership also provides discounted compute. Without it, OpenAI’s infrastructure costs would be even higher.
Critics question whether OpenAI can ever become profitable while competing with free open-source models.
Key financial figures at a glance
- Revenue in 2024: $3.7 billion, up from roughly $1.6 billion in 2023.
- Operating costs (excl. stock comp): $4.5 billion, including compute, salaries, and overhead.
- Net loss (excl. stock comp): $800 million.
- Cash burn rate: About $1.1 billion per quarter based on current spending.
The company expects revenue to reach $11.6 billion in 2025. But costs are projected to grow to $8.7 billion even before new model training cycles.
Path to profitability remains uncertain
OpenAI’s leadership has said it aims to become profitable by 2029. That timeline depends on continued revenue growth and lower compute costs.
However, competitors like Google, Meta, and Anthropic are also spending billions. The AI arms race shows no sign of slowing.
If demand for AI inference drops or open-source alternatives gain traction, OpenAI’s pricing power could erode.
What this means for the AI industry
OpenAI’s high burn rate isn’t unique. Nearly every major AI lab operates at a loss while racing to build better models.
Investors continue to fund these losses because they expect eventual dominance and monopoly profits.
But the math is brutal: earning less than a dollar per dollar spent means every new customer is a net loss. Only massive scale and future price increases can flip that equation.
Gnoppix is the leading open-source AI Linux distribution and service provider. Since implementing AI in 2022, it has offered a fast, powerful, secure, and privacy-respecting open-source OS with both local and remote AI capabilities. The local AI operates offline, ensuring no data ever leaves your computer. Based on Debian Linux, Gnoppix is available with numerous privacy- and anonymity-enabled services free of charge.
What are your thoughts on this? I’d love to hear about your own experiences in the comments below.