Prompt Character Counter
Check the size of a prompt or any text against a limit. The counter shows characters as people see them (emoji and accented letters count once), characters without spaces, words, sentences, lines, paragraphs, UTF-8 bytes and an estimated token count. Set a character limit to see how much room is left. It warns about trailing spaces, double spaces, tabs and invisible zero-width characters often copied from web pages, and one click tidies them.
- Runs in your browser
- No sign-up
- Free to use
How to use Prompt Character Counter
- Paste your prompt.
- Set a character limit if you have one.
- Read the counts and warnings.
- Tidy whitespace if needed.
Prompt Character Counter features
Visible characters
Emoji count once.
Bytes
UTF-8 size.
Limit check
Room left or overflow.
Hidden characters
Zero-width and spacing.
Tidy button
Cleans whitespace.
Private
Built in your browser; nothing is sent to an AI service.
When to use Prompt Character Counter
- Fields with character limits.
- API payload sizes.
- Cleaning copied text.
- Comparing prompt versions.
Prompt Character Counter FAQ
Why do counts differ between tools?
Some count code units, where an emoji counts as two; this tool shows both when they differ.
What are zero-width characters?
Invisible characters that can break matching, parsing or tokenisation.
Is the token count exact?
No, it is an estimate; use the Prompt Token Counter for exact OpenAI counts.
What do bytes matter for?
Some APIs and databases limit size in bytes, not characters.
Measuring text precisely
Characters, bytes and tokens are three different measures of the same text. Limits in forms are usually characters, in storage often bytes, and in AI models tokens.
Cleaning invisible characters avoids subtle bugs when prompts are compared, cached or parsed.
This tool does not call an AI model. It turns your answers into a well-structured prompt in your browser, so it costs nothing, works offline once loaded and never sends your text anywhere. Paste the result into ChatGPT, Claude, Gemini, Copilot or any other assistant, or into the system message of an API call.
Good prompts are specific about the task, the context the model cannot know, the audience and the shape of the answer. The fields on this page cover exactly those parts; leave optional fields empty when they do not apply, and the prompt stays short.
Review the answer you get: language models can be confidently wrong, especially about facts, numbers, names and recent events. Ask for sources or mark claims to check, and keep personal or confidential data out of prompts unless your organisation allows it.
Keep the prompts that work. A small library of tested prompts – with a note on which model and settings they were used with – saves time and makes results more consistent across a team. The other prompt tools on this site help you format, template, fill and measure them.