OpenAI burned through $1.22 per dollar earned even after stripping out stock-based compensation

OpenAI’s $1.22 Loss Per Dollar Earned: What This Means for the Future of AI and How It Will Be Used

In a stark reality check for the artificial intelligence industry, a newly revealed financial picture of OpenAI shows the company burned through $1.22 for every dollar it earned in the past year — and that’s after stripping out the cost of stock-based compensation. This news, emerging from OpenAI’s internal financial data and reported by The Decoder on May 22, 2026, shines a bright light on the enormous cost of pushing the boundaries of AI.

So what does this mean for companies, developers, and everyday users who are betting big on AI? It’s a story about the price of progress, the business models that might survive, and how this technology will ultimately be used. Let’s break down the numbers, explore the trends, and look ahead to what a more sustainable AI future could look like.

The Numbers That Demand Attention

To understand the scale of OpenAI’s spending, consider this: For every $1.00 in revenue the company collected, it spent $2.22 in total operating costs. After removing stock-based compensation — a non-cash expense — the ratio still stood at $1.22 spent per dollar earned. That means even in a more conservative view, OpenAI is still losing 22 cents on every dollar of sales, and the real cash burn is substantially higher.

These figures come from the most recent financial disclosures and paint a clear picture: building and operating cutting-edge AI models is astonishingly expensive. The costs go toward massive data centers, specialized chips (like GPUs), high-bandwidth networking, electricity, top-tier research talent, and the cloud infrastructure needed to serve millions of users worldwide. At the same time, OpenAI’s revenue has grown, but not enough to catch up to this breakneck spending.

This isn’t a sign of failure — it’s a signal of the investment required to lead in AI. But for the entire ecosystem, it raises an urgent question: How will AI become a profitable, sustainable industry?

What This Spends on the Future of AI Business Models

For years, many tech companies have operated on a “grow first, profit later” model. OpenAI’s situation takes that to an extreme. The company is spending to advance the frontier of artificial general intelligence (AGI), and its investors — including Microsoft — appear willing to stomach the losses for now. But there are limits. The $1.22 burn rate suggests that even a giant like OpenAI must eventually turn its technology into a self-sustaining business, or face tough choices.

Three Likely Pathways to Profitability

Higher Prices for Premium Access: One immediate way to close the gap is to raise prices. OpenAI has already introduced tiered subscription models for ChatGPT and API access. Expect future price increases, especially for enterprise customers who rely on the highest-performance models. In a world where AI is a utility, the cost of that utility will climb.
Focus on Enterprise and Industry-Specific Solutions: Rather than serving everyone at a low margin, OpenAI may double down on high-value verticals. This includes healthcare, finance, legal, and other sectors where AI can save companies enormous sums of money. Custom models, fine-tuning services, and dedicated support will command premium pricing.
Aggressive Cost Cutting Through Efficiency: Advances in chip design, model compression, and smaller, more efficient models (like “distilled” versions) can lower the cost per query. OpenAI is already investing in custom AI chips and exploring ways to reduce the computational load. If they succeed, they could drive down the $1.22 cost structure significantly.

For business leaders, the takeaway is clear: AI services are not cheap to produce, and won’t remain cheap to consume forever. The era of free or heavily subsidized AI tools may be winding down. Organizations that are building their operations on top of AI should prepare for a world where costs are more variable and potentially higher than expected.

Practical Implications for Businesses and Society

If OpenAI is burning through $1.22 for every dollar, it’s a warning light for the entire ecosystem. Many startups and developers rely on OpenAI’s APIs to power their own products. When the underlying provider is losing money, those prices will eventually go up, or the provider will change its terms. This creates risk for anyone building a business on top of someone else’s AI platform.

Here are the most important practical implications:

Diversify Your AI Suppliers: Don’t put all your eggs in one basket. Consider using multiple models from different providers or even open-source models that you can host yourself. This gives you bargaining power and resilience against price hikes.
Plan for Variable Costs: AI usage costs can spike unpredictably. Build your budgets with a buffer, and monitor your usage closely. Consider caching common queries or using smaller models for routine tasks to keep expenses under control.
Invest in AI Efficiency: The best way to manage AI costs is to use less of it. That sounds counterintuitive, but it’s true: train your teams to write better prompts, use smaller models where possible, and batch requests to reduce overhead.
Prepare for Consolidation: Not every AI company will survive this financial strain. We may see a wave of mergers, acquisitions, or even shutdowns. Bet on stable partners with strong balance sheets.

For society, this financial stress on AI leaders has a broader implication: the gap between those who can afford AI and those who cannot may widen. If AI services become too expensive, smaller businesses, educational institutions, and nonprofits may lose access to cutting-edge tools. This could concentrate AI power in the hands of a few wealthy corporations, which is exactly the opposite of the democratizing vision many technologists promote.

Governments and policymakers should take note. If the dominant AI platform is bleeding money, that’s not a stable foundation for a nation’s digital infrastructure. Supporting open-source alternatives, funding research into more efficient models, and creating incentives for affordable access are all smart moves.

The Path Forward: How AI Will Be Used

Despite the financial headwinds, the demand for AI is real and growing. The question is how will we use it when the cost structure forces efficiency? Here are the key trends that will shape the next few years:

From General to Specialized

Instead of relying on one giant, all-purpose AI model for everything, we’ll see a shift toward specialized models trained for specific tasks. A legal AI that knows contracts, a medical AI that knows disease patterns, and a customer service AI that knows your product. These smaller, focused models cost less to run and can be more accurate. OpenAI and others are already moving in this direction, but the financial pressure will accelerate it.

More Open Source, More Competition

The high cost of proprietary models like GPT-4 and GPT-5 is creating a huge incentive for open-source alternatives. Models like Llama from Meta, Mistral, and others are already closing the performance gap. As these open models improve, businesses will have more options to run AI in-house without paying per-token fees. This could put downward pressure on prices and drive innovation in efficiency.

AI as a Commodity, Not a Luxury

If costs come down through better hardware and software, AI could become a low-cost utility — like cloud computing or electricity. But that outcome is not guaranteed. The $1.22 burn rate shows we’re still in the “luxury investment” phase. To get to commodity pricing, we need breakthroughs in chip design (like Apple and Google’s custom chips) and more efficient model architectures (like mixture of experts or sparsity).

Human-in-the-Loop Becomes the Standard

When AI is expensive, you don’t use it for everything. Smart organizations will use AI as an assistant that handles the most valuable or repetitive tasks, while humans still make the final decisions. This “human-in-the-loop” model not only saves money but also improves quality and reduces risk. Expect to see more tools that combine AI suggestions with human oversight.

A Reality Check for the AI Hype Cycle

The fact that OpenAI is burning through $1.22 per dollar earned should temper some of the wilder claims about AI’s immediate impact. Yes, AI is transformative. Yes, it will change industries. But the path to profitability is long and expensive. Investors, executives, and policymakers need to plan for a multi-year journey, not a quick win.

At the same time, this isn’t a reason to panic. OpenAI’s financial position is strong enough to weather this burn rate for a while, especially with deep-pocketed backers. But it’s a reminder that even the most advanced technology must eventually pay its own way. The companies that succeed will be the ones that push toward efficiency as hard as they push for performance.

The future of AI won’t be determined by the most powerful model alone. It will be shaped by economics. The model that balances power with cost will win. And that means the next great breakthrough might not be a smarter AI — it could be a cheaper one.

TLDR: OpenAI’s financial data shows it spent $1.22 for every dollar earned, even excluding stock-based compensation. This underscores the enormous cost of building state-of-the-art AI. For businesses, this means preparing for higher prices, diversifying AI providers, and focusing on efficiency. The future of AI will be driven by specialized models, open-source alternatives, and a push toward sustainable economics. The race is no longer just about smarter AI — it’s about making AI affordable enough for widespread use.