In the field of artificial intelligence, OpenAI has undoubtedly been a focal point of attention. In the first seven months of 2025, OpenAI has demonstrated impressive growth, yet behind this brilliance lie several serious challenges.
Revenue Surge and Strong Growth Momentum
According to The Information and several other authoritative media outlets, OpenAI’s revenue has doubled in the first seven months of 2025, with an annual recurring revenue (ARR) reaching $12 billion, far exceeding the approximately $4 billion achieved in 2024. This year, OpenAI is on track to surpass its previously set annual target of $12.7 billion. Currently, OpenAI’s monthly revenue is about $1 billion, a significant leap from the $500 million in January.
This rapid revenue growth is mainly driven by the dual explosion of enterprise services and individual subscriptions. On the enterprise side, the number of paying corporate users has surged from 2 million in February 2025 to over 3 million by June, spanning use cases in programming, data analysis, and customer service. The enterprise version of ChatGPT, with customized features like Deep Research report generation and a 10%-20% discount, has successfully attracted a large number of traditional industry clients. The subscription revenue from the enterprise version now accounts for 21% of total revenue, becoming a key growth driver.
On the consumer side, the number of individual subscribers continues to expand. While the weekly active users of ChatGPT fell to 700 million in July, the number had reached 800 million in April 2025 (according to DemandSage), a doubling from the beginning of the year. Among individual subscribers, 55% are paying for ChatGPT Plus, with a growing willingness to pay. Recent features such as spreadsheet editing and presentation generation directly compete with Microsoft’s Copilot and Google Workspace, further solidifying OpenAI’s position in the consumer market.
Expanding User Base and Rising Influence
User growth is a key factor driving OpenAI’s revenue increase. The ChatGPT product currently has approximately 700 million weekly active users, covering both consumers and enterprise clients, up from 500 million in March 2025. This vast user base not only brings direct economic benefits to OpenAI but also boosts its influence and voice in the AI field.
OpenAI continues to attract more users by constantly optimizing its product features. For instance, the introduction of a 1 million-token context window has strengthened the coding capabilities of the GPT-4.1 series, which is highly attractive to the developer community. Additionally, its intelligent agent products, such as the PhD-level research agents, despite costing $20,000 per month, are targeting data-intensive industries such as finance and healthcare, further expanding its user base.
Competition and Cost Challenges
- Growing Competitive Pressure
In the AI field, OpenAI is not the only player; it faces intense competition from all sides. Traditional giants like Google and Meta are accelerating their efforts. Google’s Gemini 2.5 Pro has captured a 31% market share in reasoning tasks, surpassing Anthropic’s Claude series; Meta’s Llama 4 is grabbing market share in the open-source sector with a low-cost strategy, despite reports of data cheating. Additionally, open-source models from China, such as DeepSeek R1 and Qwen 3, are nearing the performance levels of top closed-source models, creating differentiated competition.
Vertical competitors are also emerging. Anthropic’s Claude model holds a 42% market share in enterprise-level programming tasks, more than twice that of OpenAI. Its annual revenue has grown nearly fourfold to $4 billion, and its valuation may reach $170 billion, posing a serious threat to OpenAI’s technological leadership. Meanwhile, the growing open-source ecosystem is disrupting OpenAI. The increasing adoption of open-source AI by enterprises, with over 75% of technical leaders planning to use more open-source tools, could divert OpenAI’s mid- and low-end customers. Open-source models like Meta’s Llama and Google’s Gemini, with their cost advantages, could become serious competitors.
- Ongoing Cost Pressures
While OpenAI enjoys revenue growth, it is also facing significant cost pressures. First, its computational costs continue to rise. OpenAI has raised its cash burn forecast for 2025 to $8 billion, and its server rental costs may exceed $14 billion. To support model training, it is collaborating with Microsoft and SoftBank to build the “Stellar Gateway” data center, with 75% of its computing power expected to rely on this project by 2030. However, in the short term, Microsoft Azure’s GPU rental costs still dominate, with nearly $4 billion spent in the first half of 2025 alone.
Additionally, human resources costs remain high. As of March 2025, OpenAI employed about 1,700 people, with labor costs (excluding stock compensation) reaching $700 million, and this is expected to rise to $2 billion by 2026. The intense competition for top AI talent has further escalated cost pressures, with several researchers recently leaving to join competitors. The high costs have delayed OpenAI’s profitability timeline. Despite rapid revenue growth, it is expected to incur a loss of $14 billion in 2026 and will need continuous financing to maintain operations. Its gross profit margin currently stands at only 41%, but it plans to increase this to 67% by 2028 through technological optimization, such as MoE architecture and economies of scale.
Conclusion
In summary, 2025 has been a year of both opportunities and challenges for OpenAI. The significant growth in revenue and user base has laid a solid foundation for its development, but intensified competition and soaring costs present serious hurdles for its future growth. Whether OpenAI can maintain its technological lead while effectively addressing competition and cost challenges to achieve sustainable development will be crucial to watch.
