JSON Token Counter
Estimate tokens and cost for JSON payloads.
With the default inputs, 0 tokens (estimated) — 0 words · 0 characters · 0 sentences. Enter your own numbers below to recompute instantly; the full step-by-step math is shown under the worked example.
Results update automatically as you type.
- Estimated tokens
- 0
- Words
- 0
- Characters
- 0
- Characters (no spaces)
- 0
- Sentences
- 0
- Input cost at this model(GPT-5)
- $0.0000
| Prose / English | 4 | 0 |
| JSON | 3.5 | 0 |
| Code | 3.3 | 0 |
| Non-Latin / other | 2 | 0 |
Estimate how many tokens a JSON payload uses and what it costs as model input. JSON is token-dense because of keys, quotes and punctuation, so budgeting it separately avoids surprises.
How this is calculated
We estimate tokens from your JSON using a JSON-tuned ratio (~3.5 characters per token — denser than prose because braces, quotes, commas and repeated keys all tokenize), then price that count as input at your chosen model. Structured payloads routinely cost 20–40% more tokens than the same information written as prose.
Is this a good result? What to do next
If a JSON payload's token count surprises you, it's usually the structure, not the data — keys and punctuation repeat on every record. For high-volume API calls that overhead is worth trimming; for a one-off it rarely matters.
Typical planning ranges
- JSON
- ~3.5 characters per token
- vs prose of same info
- ~20–40% more tokens
- Deeply nested / verbose keys
- even denser
Ranges are typical planning figures to sanity-check your result, not authoritative benchmarks. Your numbers will vary with use case, volume, and vendor.
How to improve this number
- Shorten or drop repeated keys; use compact arrays over objects where possible.
- Strip insignificant whitespace before sending.
- Send only the fields the model actually needs.
Common mistakes
- Estimating JSON at the prose ratio and under-budgeting.
- Sending pretty-printed JSON with unnecessary indentation.
When to use a different approach
For source code use the code token counter; for plain text use the token counter.
Worked example (defaults)
With the default inputs above, here is the result:
- Estimated tokens
- 0
- Words
- 0
- Characters
- 0
- Characters (no spaces)
- 0
- Sentences
- 0
- Input cost at this model(GPT-5)
- $0.0000
| Prose / English | 4 | 0 |
| JSON | 3.5 | 0 |
| Code | 3.3 | 0 |
| Non-Latin / other | 2 | 0 |
- Heuristic: tokens ~= characters / 3.5
- 0 chars -> ~0 tokens
Sources & references
Frequently asked questions
Why does JSON use more tokens?+
Braces, quotes, commas and repeated keys all become tokens, so structured JSON is denser than plain prose of the same length.
Related calculators
Answering a real question?
This calculator powers these problem-solving guides: