docs: update tiktoken readme (#40)

README.md

@@ -32,6 +32,7 @@ Next-generation AI gateway, using OpenAI as the protocol entry point.
 - Model RPM TPM limits
 - Think model support `<think>` split to `reasoning_content`
 - Prompt Token Cache billing support
+- Inline tiktoken, no need to download tiktoken file
 
 ## Deploy
 

README.zh.md

@@ -33,6 +33,7 @@
 - 模型 RPM TPM 限制
 - Think 模型支持 `<think>` 切分到 `reasoning_content`
 - 提示词缓存计费支持
+- 内敛分词器，无需额外下载 tiktoken 文件
 
 ## 部署