Here’s the translation:
Dear Mr. Altman,
GPT’s capabilities are already very powerful. However, Azure has severely restricted customers’ use of OpenAI models. On Azure, the stability and latency of public resources for GPT cannot be guaranteed, making them unsuitable for production-level use. Companies are forced to purchase Microsoft’s Provisioned Throughput Units (PTUs) at high prices, which are expensive and unaffordable for most businesses. Please consider providing more stable API services for small and medium-sized enterprises. On AWS, Claude’s API stability is essentially the same for both large and small customers.
Sincerely