← Back to Tools

🔢 Token Counter

Estimate how many tokens your text uses across major LLM tokenizers. Runs entirely in your browser — no text is sent anywhere.

Characters: 0  ·  Words: 0  ·  Lines: 0

🤖

GPT-4 / GPT-3.5

cl100k_base (tiktoken approximation)

0

tokens

Context window (GPT-4o) 128,000
0% 128,000 remaining
🧠

Claude 3.5 / Claude 4

Anthropic tokenizer (approximation)

0

tokens

Context window (Claude 4 Opus) 200,000
0% 200,000 remaining

Context window reference

Model Context % Used Remaining

How token counting works

This tool uses a rule-based approximation of the cl100k_base tokenizer (used by GPT-4 and GPT-3.5-turbo) and a similar heuristic for Claude's tokenizer. The actual token count from the official API may differ by ±5–10% for most text.

Token counts matter because LLMs have a fixed context window — sending more tokens than the limit will cause the model to truncate your message or return an error.

Note: Tokens are not characters. In English, 1 token ≈ 4 characters or ¾ of a word. Code and non-English text typically uses more tokens per character.