LLM Token & Word Counter
Get a quick estimate of prompt tokens, words, and characters before sending text to a language model.
Token count is a rough character-based estimate, not a model-specific tokenizer result. Exact counts vary by model and input.
How Token Counting Works
Language models read tokens: pieces of text that may be whole words, word fragments, punctuation, or whitespace. This page estimates token count with a simple character-based calculation (one token per four characters). It is useful for a quick size check, but it does not run the tokenizer for GPT, Claude, or another specific model.
Why the estimate can differ
Tokenizers split text differently. A common English word may use one token, while an uncommon name, source-code symbol, URL, or non-English phrase can use several. The same prompt can therefore have different token counts across models. For example, this counter treats a 400-character input as about 100 tokens regardless of its contents; a model tokenizer may return a different number.
What is included in the count
The estimate covers only the text pasted into the box. It does not add system instructions, chat-role formatting, tool definitions, image or audio payloads, or the model’s response. Use the model provider’s tokenizer or API usage fields when you need an exact context-window or billing figure. This page makes no price estimate because rates vary by model and can change.
What this token counter measures
It counts words and characters in the text box and estimates tokens at roughly one token per four characters. Use it to compare prompt drafts or spot a large input before sending it to a model; do not treat the estimate as an exact model or billing count.
How to use LLM Token & Word Counter
- Paste the prompt text you want to size into the input box.
- Read the live word, character, and estimated-token totals.
- Edit the prompt and compare the totals. For exact model usage, check the provider’s tokenizer or API response.
LLM Token & Word Counter features
- Live word and character counts update as you type.
- A model-independent token estimate gives a quick size comparison.
- Text analysis runs in the browser; the page does not calculate model-specific prices.