How many words is 1 million tokens?

One million tokens is approximately 750,000 English words, using the standard ratio of 1 token ≈ 0.75 words.

How do you convert tokens to words?

Multiply the number of tokens by 0.75 to get the approximate word count. For example, 100,000 tokens × 0.75 = 75,000 words.

How many tokens is 1 word?

One English word is approximately 1.33 tokens on average. This varies by word length and tokenizer.

How many pages is 1 million tokens?

One million tokens is about 1,500 single-spaced pages, at roughly 500 words per page.

Are tokens the same as words?

No. A token is a subword unit used by AI models — roughly ¾ of an English word or about 4 characters. Common short words like "the" are one token, while longer words may be split into multiple tokens.

Tokens to Words Converter

Convert AI tokens to words, pages, novels, and more — instantly. Works for GPT, Claude, Gemini, and all major LLMs.

Tokens

Words

750,000words

4,000,000characters

1,500pages

8.3novels

10,000paragraphs

50,000sentences

100,000lines of code

83hours of speech

What would this cost?

GPT-5.6 Sol$5.00 in$30.00 out GPT-5.5 Pro$30.00 in$180.00 out Claude Fable 5$10.00 in$50.00 out GPT-5.6 Terra$2.50 in$15.00 out GPT-5.5$5.00 in$30.00 out Claude Opus 4.8$5.00 in$25.00 out

Based on the standard ratio of 1 token ≈ 0.75 English words ≈ 4 characters. Actual counts vary by language, formatting, and tokenizer.

Common token conversions

Tokens	Words	Characters	Paragraphs	Sentences	Pages	Equivalent to
1,000	750	4,000	10	50	1.5	A short blog post
10,000	7,500	40,000	100	500	15	A long article or report
100,000	75,000	400,000	1,000	5,000	150	A short novel
128,000	96,000	512,000	1,280	6,400	192	GPT-4 context window
200,000	150,000	800,000	2,000	10,000	300	Claude Haiku context window
500,000	375,000	2,000,000	5,000	25,000	750	~4 novels
1,000,000	750,000	4,000,000	10,000	50,000	1,500	8 novels or 69% of Harry Potter
1,050,000	787,500	4,200,000	10,500	52,500	1,575	GPT-5.6 full context window
10,000,000	7,500,000	40,000,000	100,000	500,000	15,000	A small library

How token-to-word conversion works

AI models don't process text as words — they use tokens, which are subword units typically 3-4 characters long. The word "hamburger" becomes two tokens ("ham" + "burger"), while common words like "the" are a single token.

The standard rule of thumb across all major AI providers is:

1 token ≈ 0.75 English words (or equivalently, 1 word ≈ 1.33 tokens)
1 token ≈ 4 characters of English text
1 token ≈ ¾ of a word — the ratio OpenAI, Anthropic, and Google all cite

This ratio holds well for standard English prose. Code, non-English languages, and heavily formatted text may have different ratios. See our full guide to what tokens are for more detail.

Frequently asked questions

How many words is 1 million tokens?: One million tokens is approximately 750,000 English words. This is roughly 8 novels, 1,500 pages, or 69% of the Harry Potter series.
How do you convert tokens to words?: Multiply the number of tokens by 0.75. For example: 100,000 tokens × 0.75 = 75,000 words. To go from words to tokens, divide by 0.75 (or multiply by 1.33).
How many tokens is 1 word?: One English word is approximately 1.33 tokens on average. Short common words like "the" or "is" are typically 1 token, while longer or unusual words may be 2-3 tokens.
Are tokens the same across different AI models?: Not exactly — GPT, Claude, and Gemini each use different tokenizers. But the ~0.75 words-per-token ratio holds reasonably well across all of them for English text.
How many pages is 1 million tokens?: About 1,500 single-spaced pages at 500 words per page — a stack of paper roughly 15 cm (6 inches) tall.