What is a token in AI?

Tokens are the basic units that AI models use to process text. A token can be a word, part of a word, or punctuation. In English, 1 token is approximately 4 characters or 0.75 words. Code typically uses more tokens due to special characters and syntax.

How much do AI API tokens cost?

AI API pricing varies by model. As of 2026: GPT-5 costs $1.25/1M input tokens, GPT-4o costs $2.50/1M, Claude Opus 4.5 costs $5.00/1M, Claude Sonnet 4 costs $3.00/1M, Gemini 2.5 Pro costs $1.25/1M, and Gemini 2.0 Flash costs $0.10/1M input tokens.

What is a context window in AI?

A context window is the maximum number of tokens an AI model can process in a single conversation. GPT-5 and GPT-4o have 128K token windows, Claude models have 200K, and Gemini models support up to 1 million tokens.

AI Token Counter & Cost Estimator

Your Text 0 characters | 0 words

Maximum 500,000 characters. All processing happens in your browser.

Total Tokens (GPT-4)

Using cl100k_base encoding

Estimated Cost (Input)

$0.00

GPT-5 pricing ($1.25/1M tokens)

Model Comparison

G GPT-5

0 $0.00 128K

G GPT-4o

0 $0.00 128K

C Claude Opus 4.5

0 $0.00 200K

C Claude Sonnet 4

0 $0.00 200K

G Gemini 2.5 Pro

0 $0.00 1M

G Gemini 2.0 Flash

0 $0.00 1M

Context Window Usage

GPT-5 / GPT-4o (128K) 0%

Claude Opus 4.5 / Sonnet 4 (200K) 0%

Gemini 2.5 Pro / 2.0 Flash (1M) 0%

About Token Counting

Tokens are the basic units that AI models use to process text. A token can be a word, part of a word, or punctuation.

English text: ~1 token per 4 characters or ~0.75 tokens per word
Code: Often more tokens due to special characters and syntax
Other languages: May use more tokens per character

Note: This tool uses OpenAI's tiktoken (cl100k_base) encoding. Claude and Gemini use different tokenizers, so counts are estimates based on typical ratios.