All tools

Token counting and prompt word writing tool

Use o200k_base or cl100k_base to count text tokens, combine system prompt words and edit scene templates. The tool does not invoke models, generate AI answers, or calculate real bills.

Select tool

Token count
Press o200k_base or cl100k_base to count text tokens, and check the Unicode code point, UTF-8 byte count and Token ID at the same time. Suitable for comparing text lengths without passing the results off as model bills.
System prompt word construction
Assemble roles, tasks, context, constraints, and output formats into clearly structured system prompts. Provides code review and translation examples, local editing and copying, without calling AI or executing prompt words.
prompt word template
Browse 18 prompt word templates, search by category and Chinese and English keywords, replace placeholders locally, edit and copy. Covers code review, debugging, writing, translation, and work planning without invoking AI.

How to choose

Estimate text length: The token count is related to the selected encoding and is not equal to the number of words; chat packages, images, and tool calls may also generate additional usage.
Organizational mission requirements: The system prompt word builder organizes text according to modules such as roles, tasks, backgrounds, constraints, and output formats. Prioritize writing clearly what is to be accomplished and how to judge the results.
Adjust scene template: The template is the starting point, replace the placeholder content and supplement the real background, and then verify the effect in the target model. The local editor does not automatically send prompt words.