Claude Opus 5 is here, Fable 5 performance, half the price

Zhishi Author | Yunpeng Editor | Mo Ying Zhixi reported on the morning of July 25th that early this morning, Anthropic's new generation flagship model Claude Opus 5 was officially released and launched online. It combines the characteristics of "thoughtful" and "active response" and is very close to the Claude Fable 5 in performance, but the price is only half. At this time, netizens commented: "If it is almost the same as Fable 5, then why do you need to spend twice the price to use Fable 5?" Back to the model, in terms of programming, Claude Opus 5 achieved SOTA in programming and knowledge work assessments such as Frontier-Bench and GDPval-AA, but it still lags behind Mythos 5 in network security tasks. Anthropic really not only boasted about the new model this time, but also saved enough face for the previously popular Fable 5 and Mythos 5. It can be said to be the "Master of Duan Shui". In terms of price, Claude Opus 5 is priced at USD 5 per million Tokens for input / USD 25 per million Tokens for output (the same as Opus 4.8). Developers can use Claude-opus-5 in the API. Fast mode runs about 2.5 times the default speed and costs 2 times the base price. This time Claude has also updated two functions that are in the testing stage: the cache of tool tip words will not be invalidated when changing the tool tip during the conversation, and automatic API downgrade processing (when encountering a request marked by a security classifier) ​​to avoid direct rejection of the request. Anthropic mentioned that Opus 5 is designed for daily high-frequency use and works more efficiently than previous models. It has become the new default model of Claude Max and is the most powerful model among Claude Pro. Regarding the release of Opus 5, netizens have been discussing it a lot, and some people also took out the "old picture": Some netizens complained about Anthropic's specialty: reset