引入新的 Anthropic 令牌计数 API

使用 Claude 时请密切关注您的成本

Anthropic has quietly released some exciting beta features over the last couple of days. One of these was the ability to process PDFs with their models, which can now understand both text and visual content within PDF documents. I’ll maybe write up something on that at a later date.

Anthropic 在过去几天悄悄发布了一些令人兴奋的测试版功能。其中之一是能够使用模型处理 PDF，该模型现在可以理解 PDF 文档中的文本和视觉内容。稍后我可能会写一些相关内容。

The other exciting beta feature, and the subject of this article, was the introduction of token counting. Crucially, you can count the tokens in user messages, PDFs and images before you send them to Claude. This is excellent news for those who like to monitor their token usage costs closely.

另一个令人兴奋的测试版功能，也是本文的主题，是令牌计数的引入。至关重要的是，您可以在将用户消息、PDF 和图像发送给 Claude 之前计算它们中的令牌。对于那些喜欢密切监控代币使用成本的人来说，这是个好消息。

According to the official announcement from Anthropic (link here),

根据 Anthropic 的官方公告（链接在此），

“The token counting endpoint accepts the same structured list of inputs for creating a message, including support for system prompts, tools, images, and PDFs. The response contains the total number of input tokens.”

“令牌计数端点接受用于创建消息的相同结构化输入列表，包括对系统提示、工具、图像和 PDF 的支持。响应包含输入令牌的总数。”

And supports the following models,

并支持以下型号，

“Claude 3.5 SonnetClaude 3.5 HaikuClaude 3 HaikuClaude 3 Opus”

“克劳德 3.5 十四行诗克劳德 3.5 俳句克劳德 3 俳句克劳德 3 作品”