Token Prices Fall as GPU Rental Costs Climb Since June
源：https://t.co/Vne0l6yByz
The margin in AI may be shifting from model providers to whoever owns compute, as inference prices fall while GPU rents rise. Builders choosing between open-weight and closed models can cut inference costs to roughly a fifth by picking open alternatives that run on older GPUs. DeepSeek's 51% price cut keeps pressure on OpenAI and Anthropic pricing across Asian markets.
Since June, the average paid token price from OpenAI has dropped 62%, with DeepSeek down 51%, Google down 29% and Anthropic down 11%, according to Ornn.
Over the same three months, Ornn tracked B200 GPU rental rates rising 50% and H200 rising 28%.Ornn calculates that the cheapest open-weight model completes a task for about 1/5 the cost of a comparable closed model and does not require the newest chips.
源：https://t.co/Vne0l6yByz
源： x.com/OrnnExchange/s… Ornn @OrnnExchange The open vs closed source race is a price war. Since June, the average paid token from OpenAI got 62% cheaper. DeepSeek: −51%. Google: −29%. Anthropic: −11%. Over the same three months, Ornn has tracked B200 rent rising 50%. H200 rose 28%. Models are getting cheaper to buy. The compute to run them is getting more expensive to rent. So where does value accrue? To whoever owns the compute. The cheapest open-weight model finishes a task for about 1/5 the cost of a comparable closed model, and they don't need the newest chips. Every GPU is a shovel when the models are free. See more at data.ornn.com . 🔗 View Quoted Tweet
This report is categorized under AITOP's "Industry" desk. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison and AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include OpenAI, DeepSeek, price war, open-source models.
Source attribution: curated from AI Will (@FinanceYF5). Cross-source corroboration: 11 independent sources have covered this development (including Blognone, The Information, Guillermo Rauch), confirming industry-wide signal strength.
