OpenAI has reached a major milestone of one billion active users while simultaneously making its AI models dramatically cheaper, cutting prices for two GPT-5.6 variants by up to 80% as of late July. The price reductions reflect efficiency gains across the company's serving systems, though OpenAI continues to lose billions annually despite its massive user base.

What's Driving OpenAI's Aggressive Price Cuts?

On July 30, OpenAI reduced prices for GPT-5.6 Luna by 80 percent and GPT-5.6 Terra by 20 percent, while keeping its flagship Sol model pricing unchanged. The Luna model now costs $0.20 per million input tokens and $1.20 per million output tokens, down from $1 and $6 respectively. Terra dropped from $2.50 and $15 to $2 and $12 per million tokens. The company attributes these cuts to efficiency improvements across its infrastructure and production systems.

OpenAI's technical team achieved these gains through multiple optimization strategies. The company used a human-led process where GPT-5.6 Sol autonomously rewrote and optimized production GPU (graphics processing unit) kernels, reducing end-to-end serving costs by 20 percent. Additionally, experiments to improve speculative decoding, a technique that accelerates token generation, increased efficiency by over 15 percent.

"Better intelligence drives broader adoption. Broader adoption supports more investment. More investment improves intelligence and efficiency," stated Sarah Friar, OpenAI's Chief Financial Officer.

Sarah Friar, Chief Financial Officer at OpenAI

How Are These Price Changes Affecting Users and Developers?

For API users and businesses relying on OpenAI's models, the pricing changes create several practical opportunities and considerations:

• Model Selection Strategy: GPT-5.6 Luna is now significantly cheaper at $0.20 and $1.20 per million tokens, making it suitable for workloads previously assigned to higher-priced models that may have required more expensive options.
• Subscription Credit Efficiency: The reduced credit consumption of Luna and Terra under Codex and ChatGPT Work subscriptions means the same subscription budget now covers substantially more output from these models without any change to subscription pricing.
• Speed vs. Cost Trade-off: OpenAI introduced Fast mode for Sol, offering up to 2.5 times the standard processing speed at double the price, allowing developers to evaluate whether faster responses justify the additional cost for their use cases.
• Flagship Model Stability: Sol pricing remains unchanged, so workloads relying on the flagship model do not benefit from cost reductions and should not expect price changes in the near term.

Despite these price cuts, subscription prices and quota budgets remain the same, meaning users on flat-rate plans benefit immediately from the efficiency gains without paying more.

The Profitability Paradox: One Billion Users, Billions in Losses

OpenAI's user growth has been extraordinary, but financial sustainability remains elusive. The company announced it now serves over one billion active users and more than two million businesses. However, the company generated $13.07 billion in revenue during 2025 while losing $21 billion, according to financial disclosures. Only about 50 million of the company's 900 million weekly users subscribe to paid plans, leaving the vast majority using free tiers.

Analysts warn that lower prices could further pressure margins as OpenAI continues massive spending on computing infrastructure. The flat-rate subscription model creates a particular challenge: heavy users can consume far more value than they pay for, making profitability difficult even with billions in revenue. OpenAI has not yet released updated 2026 financial figures, leaving questions about whether the price cuts will narrow or widen losses as adoption increases.

OpenAI's Broader Push to Support Academic Research

Beyond commercial pricing changes, OpenAI is investing heavily in academic adoption. The company introduced ChatGPT for Academic Researchers, a program offering 100,000 academic researchers at selected institutions free access to ChatGPT, ChatGPT Work, Codex, and frontier models including the GPT-5.6 family through 2027. The program began with 10,000 researchers this summer, with access already available at institutions such as the Institute for Advanced Study and École normale supérieure.

Researchers can invite up to four collaborators from their institution and access business-grade privacy and security protections, with data not used to train OpenAI's models by default. The program includes over 75 life science skills spanning genetics, genomics, sequencing, single-cell analysis, protein modeling, and drug discovery, along with connectors providing access to scientific literature, public genomic and clinical databases, satellite imagery, and computational notebooks.

This academic initiative is part of OpenAI's broader commitment of more than $250 million through 2027 to support external scientific research and discovery. The investment includes NextGenAI, a $50 million initiative supporting research institutions, and collaboration with the Department of Energy's Genesis Mission to bring frontier AI to researchers at national laboratories and universities.

Currently, roughly 1.3 million people use ChatGPT weekly for advanced science and mathematics, with AI moving from occasional use on isolated problems to a regular part of mathematical research. A growing number of academic papers now acknowledge ChatGPT's contribution, reflecting how rapidly researchers are adopting these tools in their work.