The 80% price cut is the tell.OpenAI and Anthropic have been trading LLM price cuts since June, and the cache read line is where the real fight is. When agents do coding work, cache reads are 96 to 99 percent of total token spend, so Anthropic cutting Fable 5.1's cache cost 75% cuts real usage cost 70 to 75 percent. OpenAI answered by halving GPT-6.1 Sol's cache price.Nobody cuts 80% over a rival.#llm #openai #anthropichttps://links.madcoolsocial.com/8EGMJsj
The 80% price cut is the tell.OpenAI and Anthropic have been trading LLM price cuts since June, and the cache read line ...
