OpenAI and Anthropic Forecast Massive Revenue Growth
Top AI companies are forecasting massive revenue growth, with internal estimates projecting $180 billion in annual recurring revenue for OpenAI and $149 billion for Anthropic by 2029. Analysis on the 20VC podcast suggests this growth reflects a belief that AI will significantly expand the total addressable market for software, which currently stands at around $700 billion worldwide.
- OpenAI's annualized revenue run rate surpassed $20 billion in 2025, a significant jump from $6 billion in 2024 and $2 billion in 2023. This growth is largely driven by enterprise adoption, with teams in over 80% of Fortune 500 companies using ChatGPT by the end of its first year. - Anthropic's revenue has also seen explosive growth, hitting a $9 billion annualized run rate at the end of 2025, up from $1 billion at the end of 2024. The company projects its revenue could reach $26 billion in 2026, with a significant portion—approximately 80%—coming from over 300,000 business customers using its API. - This market expansion is attracting massive venture capital investment into AI agents. According to PitchBook, 30% of all venture funding in 2025 was expected to go to AI-native startups. Companies are building agentic systems for specialized industries, such as financial crime compliance (WorkFusion), healthcare automation (EliseAI), and even automated software engineering (Cognition AI). - Venture capital firm Sequoia Capital has stated that the value in the AI boom will primarily be captured at the application layer, not the foundational model layer. Their analysis suggests AI is attacking both the software and services markets, creating a market opportunity potentially 10 times larger than cloud computing. - In the real estate sector, AI agents are being integrated into consumer platforms and back-office operations. Zillow is using AI to create CRM tools for agents that summarize calls and draft follow-up messages, while Rocket Companies launched an app within ChatGPT to facilitate conversational home searches. - Startups are also building vertical-specific agents to solve industry pain points. For example, MARC, a Dublin-based startup, raised $1 million to build AI agents that automatically extract key terms from thousands of vendor contracts for real estate asset managers. - The growth is not without substantial costs; OpenAI's spending on computing power is immense, with a reported annual burn rate of $17 billion. The company's spending is expected to continue, with internal forecasts projecting operating losses could reach approximately $74 billion by 2028 before potentially turning a profit in 2030. - Anthropic, while also spending heavily on training costs, is projected to reach cash-flow positive status sooner than OpenAI, with current forecasts pointing to 2028. This is attributed to a strong focus on B2B sales of its Claude models through APIs on platforms like Google Vertex AI and AWS Bedrock.
