$200 Subscription Could Burn $14,000 in Computing Power, Major AI Models Face Cost Crunch

Industry analysis firm SemiAnalysis recently tested the subscription plans of OpenAI and Anthropic in real-world scenarios. The findings reveal that behind the seemingly low fixed monthly fees, there is a substantial subsidy gap for computing power that large model providers must cover.
Testers purchased various subscription tiers from both companies and continuously ran resource-intensive tasks like long-term programming and autonomous agents until hitting weekly usage caps. They then calculated the theoretical costs based on publicly available API pricing, and the results were startling.
Computing Power Subsidies Pushed to Their Limits
The calculations showed that if a user fully exploits the OpenAI "ChatGPT Pro 20x" subscription at $200 per month, the equivalent API billing could reach roughly $14,000. Similarly, the Anthropic "Claude Max 20x" plan at the same price could theoretically incur token costs up to $8,000 under extreme use.
This means a small number of heavy users can eat into the already thin profit margins of subscription models, pushing providers into significant losses. For ChatGPT Plus, a $20 entry-level plan, OpenAI begins losing money once a user's utilization exceeds 11.4%.
Enterprise Task Routing Emerges as a Key Strategy
In this environment, agent systems relying on multi-step and autonomous tool calls are escalating cost pressures, with token consumption reaching thousands of times that of traditional single-turn chats. Major companies like Microsoft, Meta, and Amazon have begun scaling back earlier practices of encouraging employees to extensively test AI, aiming to curb rapidly growing internal expenses.
To tackle high computing costs, more enterprises are adopting a refined task distribution approach: complex problems go to top-tier models, while routine office work is handled by cheaper or open-source models. This task routing method can cut overall AI costs by up to 95%, but it also forces large model providers to struggle with balancing user experience and heavy infrastructure investments.
Related article
U.S. Stocks Hit Historic Milestone as AI and Aerospace Giants Prepare for Trillion-Dollar Debut
Elon Musk, Sam Altman, and Dario Amodei, three titans of the technology sector, are advancing toward initial public offerings for their respective ventures. With SpaceX, OpenAI, and Anthropic—three industry behemoths nearing trillion-dollar valuation
Swedish AI Startup Lovable Eyes $13.2 Billion Valuation After Major Funding Round
As AI-driven coding tools gain traction, Swedish startup Lovable has secured a major funding round. The company aims to raise $3 billion, potentially boosting its valuation to $13.2 billion—double the $6.6 billion recorded last December. Menlo Ventur
Google Tests Remy AI Agent for Gemini as Focus Shifts to User Control
According to Business Insider, Google is testing Remy, a new AI personal agent for Gemini. This tool aims to execute tasks on behalf of users, streamlining both professional workflows and daily routines.Currently, Remy is undergoing testing in an int
Related Special Topic Recommendations
Comments (0)
0/500

Industry analysis firm SemiAnalysis recently tested the subscription plans of OpenAI and Anthropic in real-world scenarios. The findings reveal that behind the seemingly low fixed monthly fees, there is a substantial subsidy gap for computing power that large model providers must cover.
Testers purchased various subscription tiers from both companies and continuously ran resource-intensive tasks like long-term programming and autonomous agents until hitting weekly usage caps. They then calculated the theoretical costs based on publicly available API pricing, and the results were startling.
Computing Power Subsidies Pushed to Their Limits
The calculations showed that if a user fully exploits the OpenAI "ChatGPT Pro 20x" subscription at $200 per month, the equivalent API billing could reach roughly $14,000. Similarly, the Anthropic "Claude Max 20x" plan at the same price could theoretically incur token costs up to $8,000 under extreme use.
This means a small number of heavy users can eat into the already thin profit margins of subscription models, pushing providers into significant losses. For ChatGPT Plus, a $20 entry-level plan, OpenAI begins losing money once a user's utilization exceeds 11.4%.
Enterprise Task Routing Emerges as a Key Strategy
In this environment, agent systems relying on multi-step and autonomous tool calls are escalating cost pressures, with token consumption reaching thousands of times that of traditional single-turn chats. Major companies like Microsoft, Meta, and Amazon have begun scaling back earlier practices of encouraging employees to extensively test AI, aiming to curb rapidly growing internal expenses.
To tackle high computing costs, more enterprises are adopting a refined task distribution approach: complex problems go to top-tier models, while routine office work is handled by cheaper or open-source models. This task routing method can cut overall AI costs by up to 95%, but it also forces large model providers to struggle with balancing user experience and heavy infrastructure investments.
U.S. Stocks Hit Historic Milestone as AI and Aerospace Giants Prepare for Trillion-Dollar Debut
Elon Musk, Sam Altman, and Dario Amodei, three titans of the technology sector, are advancing toward initial public offerings for their respective ventures. With SpaceX, OpenAI, and Anthropic—three industry behemoths nearing trillion-dollar valuation
Swedish AI Startup Lovable Eyes $13.2 Billion Valuation After Major Funding Round
As AI-driven coding tools gain traction, Swedish startup Lovable has secured a major funding round. The company aims to raise $3 billion, potentially boosting its valuation to $13.2 billion—double the $6.6 billion recorded last December. Menlo Ventur





Home






