Kimi Open Platform to Launch ContextCaching Function Internal Test

TapTechNews June 19th-reported that the Dark Side of the Moon officially announced that the ContextCaching function of the Kimi open platform will start internal testing, and it will support long-text large models and can realize the context caching function.

According to the introduction, ContextCaching (context caching) is an advanced function provided by the Kimi open platform, which can reduce the cost when users request the same content by caching repeated Token contents, and the principle is as follows:

The official said that ContextCaching can improve the interface response speed of the API (or the first character return speed). In the prompt scenarios with large scale and high repetition, the more benefits the ContextCaching function brings.

ContextCaching is suitable for situations with frequent requests and repeated references to a large amount of initial context. By reusing the cached content, it can improve efficiency and reduce costs. The applicable business scenarios are as follows:

A QABot that provides a large amount of preset content, such as the KimiAPI little assistant.

Frequent inquiries for a fixed set of documents, such as the information disclosure Q&A tool for listed companies.

Periodic analysis of static code libraries or knowledge bases, such as various CopilotAgents.

Extremely popular AI applications with huge instantaneous traffic, such as Honghong Simulator, LLM Riddles.

Agent-type applications with complex interaction rules, such as What's Worth Buying Kimi+.

The official will later release the scene best practice/billing plan/technical documentation of the ContextCaching function, and TapTechNews will keep paying attention and bring the relevant reports as soon as possible.