OpenAI has significantly improved its operational efficiency in artificial intelligence (AI), with its computing margin reaching 70% in October, up from 52% at the end of last year and double the 35% recorded in January 2024. This internal metric measures the share of revenue after deducting costs for running models for enterprise and consumer users. Despite ongoing unprofitability due to high computing costs, OpenAI is achieving higher returns on server investments supporting ChatGPT subscriptions and enterprise model licensing.
Two key factors drive OpenAI's margin expansion: declining annual computing lease costs and optimized AI models for greater efficiency. Additionally, higher-priced subscription tiers have boosted revenue from select customer segments. In contrast, rival Anthropic reported a -90% computing margin last year, though projections suggest it could rise to 53% by year-end and potentially 68% next year. However, when factoring in free-user model costs and training expenses, Anthropic's overall server efficiency may surpass OpenAI's.
With most ChatGPT users on free plans, OpenAI must monetize this segment through ads or shopping commissions to close the efficiency gap. The company is also aggressively marketing industry-specific enterprise solutions, competing with Alphabet (GOOGL.US) and Anthropic in sectors like finance and education.
OpenAI's improved margins come as it seeks $100 billion in new funding, which could value the company at $830 billion if fully raised. The funding round is expected to conclude by Q1 next year.
Meanwhile, CEO Sam Altman has declared a "code red," pausing side projects like Sora video generation for eight weeks to focus on ChatGPT enhancements. This reflects internal tensions between consumer growth and AGI research priorities as competition intensifies. Alphabet's Gemini 3 model recently outperformed GPT-5.1 in benchmarks, scoring record highs in reasoning (37.5% vs GPT-5 Pro's 31.64%) and academic tests (91.9% in GPQA Diamond). Anthropic strengthened its position through a $350 billion partnership with Microsoft (MSFT.US) and Nvidia (NVDA.US).
While ChatGPT maintains dominance with 90% mobile chatbot session share (vs Gemini's 4%) and 8+ daily opens per user (vs Gemini's 2.5), Alphabet's Gemini app now claims 650 million monthly active users compared to ChatGPT's 800 million weekly actives. Analysts warn that Alphabet's scale and search dominance could further erode OpenAI's lead.
OpenAI's financial sustainability faces pressure as it delays monetization projects like AI-powered ads and shopping assistants to prioritize ChatGPT. Despite 200% revenue growth to $4.3 billion H1 2025, net losses ballooned to $13.5 billion from $3.1 billion YoY. With only 5% of its 800M+ MAUs paying $20 monthly subscriptions, HSBC projects OpenAI won't achieve profitability by 2030, facing a $207 billion cumulative cash shortfall amid $792 billion projected computing costs through 2030.
Comments