"OpenAI says its revenue surged to $20 billion in 2025, showing massive growth even as rivals like Anthropic and Google gain ground in AI markets."
OpenAI crosses a major revenue milestone
OpenAI has revealed that its annualized revenue reached $20 billion in 2025, a stunning jump of 233 percent compared to the previous year. For anyone following the AI industry, this number is eye catching and signals just how fast demand for generative AI tools has grown.
According to OpenAI, revenue climbed from $2 billion in 2023 to $6 billion in 2024, before exploding to $20 billion in 2025. In just two years, the company multiplied its revenue nearly ten times. CFO Sarah Friar described it as growth on a scale rarely seen in modern tech.
How did OpenAI grow so fast?
The simple answer is usage. More businesses, developers, and everyday users are relying on AI for writing, coding, customer support, research, and creative work. Products like ChatGPT and OpenAI APIs are now deeply embedded in many workflows.
But there is another important factor behind the scenes: computing power.
Compute expansion and revenue growth go together
OpenAI says its revenue curve closely followed its compute expansion. The company scaled its infrastructure aggressively over the past three years.
| Year | Compute Capacity | Revenue |
|---|---|---|
| 2023 | 0.2 GW | $2B |
| 2024 | 0.6 GW | $6B |
| 2025 | 1.9 GW | $20B |
By 2025, OpenAI was running close to 1.9 gigawatts of compute power. That is roughly equal to the electricity usage of around two million households. The company says compute grew three times year over year, or 9.5 times compared to 2023, and revenue followed the same path.
The massive cost of scaling AI
Growing this fast is not cheap. Training and running large language models requires enormous capital. Reports suggest OpenAI has been exploring fundraising on an unprecedented scale.
According to the Wall Street Journal, OpenAI was seeking to raise up to $100 billion at a valuation near $830 billion to continue expanding its infrastructure. Around the same time, SoftBank reportedly completed a $40 billion investment, highlighting how serious investors are about AI infrastructure.
AI growth today is limited less by demand and more by how fast companies can build compute.
Competition is catching up fast
While OpenAI dominates headlines, the competitive picture is changing quickly.
Enterprise AI market share shifts
In the enterprise large language model API market, OpenAI is no longer the clear leader. Data from Menlo Ventures shows OpenAI market share dropping from about 50 percent in late 2023 to roughly 25 percent by mid 2025.
- Anthropic leads with around 32 percent market share
- Google follows with about 20 percent
In coding related tasks, the shift is even more dramatic. Anthropic's Claude reportedly holds about 42 percent market share, while OpenAI sits near 21 percent.
Consumer traffic tells a similar story
On the consumer side, ChatGPT still leads, but its dominance is shrinking. SimilarWeb data shows ChatGPT market share falling from nearly 87 percent at the start of 2025 to about 64.5 percent by early 2026.
During the same period, Google Gemini crossed the 20 percent mark for the first time, reaching roughly 21.5 percent. This suggests users are increasingly trying multiple AI tools rather than relying on a single platform.
Ads arrive on ChatGPT free plans
To diversify revenue, OpenAI has announced a significant strategic change. The company will begin testing advertisements on the free tier of ChatGPT and on its new $8 per month Go subscription plan for users in the United States.
Paid plans like Plus, Pro, and Enterprise will remain ad free. CEO Sam Altman acknowledged that ads were once seen as a last resort, but growing competition and infrastructure costs are clearly reshaping priorities.
What this means for the future of AI
OpenAI hitting $20 billion in revenue proves that AI is no longer experimental. It is a core part of the global tech economy. At the same time, the rise of Anthropic and Google shows that leadership in AI can shift quickly.
The next phase of competition will likely focus on efficiency, enterprise trust, pricing, and real world usefulness rather than just raw model size.
Frequently Asked Questions
Is OpenAI profitable at $20 billion revenue?
Not necessarily. OpenAI spends heavily on compute, research, and infrastructure, so high revenue does not automatically mean high profit.
Why is compute so important for AI companies?
Compute determines how powerful models can be and how many users they can serve. More compute usually means better performance and higher costs.
Will ads change the ChatGPT experience?
For free users, ads may become part of the experience. Paid users on higher plans will not see ads, at least for now.
