Just like gold and oil, we’ll soon be able to trade AI token futures

The emergence of AI tokens is poised to reshape the financial landscape, with significant interest from major financial institutions. According to reports, the Shanghai Futures Exchange in China is in the process of creating a derivatives market specifically for AI tokens. This initiative coincides with separate efforts by leading exchanges, including CME Group and the Intercontinental Exchange, to roll out futures contracts related to GPU rentals. As the demand for graphics processing units (GPUs) continues to grow, the market for GPU rentals is evolving. AI Mining Co. has been tracking daily pricing data across various platforms, revealing that the median rental prices for Nvidia H100 GPUs fluctuate between $1.40 and $4.27 per hour. Meanwhile, prices for the newer H200 models range from $2.34 to $5 per hour. Over the past week, the average rental price for H100 GPUs has settled between $2.79 and $3.33. While GPU rental markets are becoming increasingly established, the infrastructure surrounding AI tokens—integral components of modern AI frameworks—remains underdeveloped. Major AI firms commonly operate with pricing structures based on tokens. For instance, OpenAI charges $5 for every million input tokens and $30 for output tokens when utilizing its latest GPT-5.5 API. Additionally, cloud providers like Amazon are beginning to introduce per-token pricing models within their services. This initiative arrives at a time when there is a significant expansion of AI infrastructure, with cloud service providers and private equity firms investing heavily in data centers. This investment surge reflects a widespread expectation that demand for GPU resources and computational power will only continue to escalate. New global neocloud companies are emerging to capture this demand, with some focusing on specialized services for inference and others challenging established cloud leaders like Oracle, AWS, and Google Cloud. The Shanghai exchange’s plans to introduce derivative products linked to AI tokens could provide a vital mechanism for businesses, investors, and data center operators to manage their computing costs effectively, reflecting the evolving dynamics of the AI industry.