Prompt Engineering

Tokens vs Characters — Why They Differ

A token is not a character. This shows the gap on a real prompt — characters, tokens, and the ratio between them — so the difference stops being abstract.

Overview

People reach for character count because it is easy, then get surprised by token bills. This loads a short prompt with no assumed response so the focus stays on the units: roughly four characters per token in English, but it swings with content and language. Tokens are the model's unit and characters are the keyboard's — counting one to predict the other is where estimates go wrong. The report makes the ratio visible instead of leaving you to guess it.

Workflow

See both counts
Characters and tokens side by side on the same text.
Read the ratio
Roughly 4 characters per token in English — but only roughly.
Watch it move
Change the text type or language and the ratio shifts.

Why This Works

Seeing both counts on one text makes the gap concrete, not theoretical
The ~4:1 English ratio is shown, not assumed
Content-type detection demonstrates why the ratio is not fixed

Best for

Understanding why token bills surprise people
Learning the rough character-to-token ratio
Anyone using character count as a token proxy

Not for

Counting characters for a platform limit — use the Character Counter
Deciding context-window fit — use the Context Window Estimator

Use cases

Understanding why token bills surprise people
Learning the rough character-to-token ratio
Anyone using character count as a token proxy

Customize This Resource

Opens this text in Token Counter. Count to get the full token and cost report — then adjust the model and assumed response length.

Open in Token Counter

Prompt Template

TOKEN COUNT REPORT

TOKEN ESTIMATE
- Estimated tokens: ~70 (range 63–77)
- Characters: 274
- Words: 45
- Detected content type: Prose
- Tokens per word (approx): 1.6
- Tokens per character (approx): 0.3

COST ESTIMATE — GPT-5
- Pricing (approximate, June 2026): input $1.25/1M tokens · output $10.00/1M tokens
- Input cost (this prompt): $0.000079–$0.000096
- Assumed response: none (input only)
- Per 1,000 calls: $0.0788–$0.0963

MODEL NOTES
- OpenAI tokenizer (o200k-class) — English averages roughly 4 characters per token.
- Same text, estimated per model: GPT-5 ~70 · Claude Opus ~73 · Claude Sonnet ~73 · Gemini Pro ~70 (a count, not a fit check — for "will it fit?" use the Context Window Estimator).

USAGE GUIDANCE
- For scale on GPT-5: a short prompt ≈ 71, a medium prompt ≈ 506, a large prompt ≈ 3,031 tokens.
- This text is closest to a short prompt.
- Output tokens usually cost several times more than input — add an assumed response above to see the fuller cost.

ESTIMATION NOTES
- A token is not a character and not a word — it is a sub-word chunk. English averages ~4 characters / ~0.75 words per token.
- Estimates vary by tokenizer: the same text tokenizes differently on GPT, Claude, and Gemini — that is why this is a range, not a single number.
- Language matters: CJK and many non-Latin scripts use more tokens per character than English.
- Code differs from prose: symbols, indentation, and punctuation push code to more tokens per character.
- These are character-based ESTIMATES, not tokenizer output. Pricing is approximate as of June 2026; providers change rates — verify before relying on a number.

Related Resources

Explore more resources that work well with this one.