500 WORDS → TOKENS

How many tokens is 500 words?

Last updated

500 English words is about 665 tokens, using the standard average of 1.33 tokens per word. That is roughly one page of single-spaced text and fits comfortably in every modern model context window.

QUICK CONVERTER

≈ Words

500

≈ Tokens

665

Estimates use ~1.33 tokens per English word. For an exact count of real text — including code, JSON, and other languages — paste it into the token counterand choose your model.

The detail

A 500-word piece — a short blog section, a college essay paragraph block, or a long email — lands around 665 tokens of input. At typical 2026 API rates that costs a fraction of a cent to send, so for single requests the cost is negligible; it only adds up at volume.

The 1.33 tokens-per-word average holds for plain English prose. Markdown formatting, bullet lists, and links push the count up slightly; heavy code blocks or tables can double it. If your 500 words include structured content, budget 700-900 tokens to be safe.

Remember that chat requests also carry the system prompt and conversation history, so the words you paste are rarely the whole bill. Use the converter below for nearby word counts, or paste real text into the analyzer for an exact count.

Word counts near 500 words

Word~Words~Tokens
250 words250333
400 words400532
500 words500665
600 words600798
750 words750998
1,000 words1,0001,330

FAQ

How many tokens is 500 words? — FAQ

Does 500 words always equal 665 tokens?

No — 665 is the English-prose average. Simple vocabulary can come in under 600 tokens, while technical writing, code, or non-English text can run well over 800. Paste the actual text into the counter for an exact number.

How much does a 500-word prompt cost?

At $1 per million input tokens, about $0.0007 per request. Even at premium rates around $5 per million, a 500-word prompt costs less than half a cent to send.

How long is a 500-word AI response in tokens?

The same math applies to output: a 500-word answer is about 665 output tokens. Output rates are typically 3-8x input rates, so long responses drive cost more than long prompts.