API Response Analyzer

Runs in browser API Testing

Paste a JSON response to get a structural report — key counts, depth, type distribution, empty and null fields, inconsistent types across array items, date/ID formats and size metrics.

Privacy: This tool runs entirely in your browser. Your input never leaves your device.

How to use API Response Analyzer

  1. Paste the JSON response body.
  2. Press Analyze.
  3. Read the findings list first, then drill into the tables (inconsistent types, formats, empties).
  4. Share the CSVs with the API owner or attach them to a bug.

API Response Analyzer features

  • Overview: root type, key counts, depth, objects/arrays/nulls, pretty vs minified size
  • Data-quality findings: nulls, empties, inconsistent types across array items, numbers/booleans as strings, mixed id and timestamp formats, duplicates, deep nesting
  • Type distribution and detected value formats (uuid, email, url, ISO dates, Unix timestamps, IP, JWT…)
  • Largest arrays, longest strings and most frequent keys
  • Runs locally; CSV downloads for every table

API Response Analyzer example

Spot inconsistent fields

Input:

{ "data": [ { "id": "usr_1001", "age": 29, "createdAt": "2025-03-01T10:15:00Z" }, { "id": 1002, "age": "31", "createdAt": 1740823200 } ] }

Output:

2 fields with inconsistent types across array items — $.data[*].id (string, number), $.data[*].age (number, string)
Timestamps mix Unix epoch numbers and ISO-8601 strings — pick one representation.

Frequently asked questions about API Response Analyzer

What issues does it detect?

Null and empty fields, fields whose type changes between array items, numbers and booleans sent as strings, mixed id formats, mixed timestamp representations, duplicate array items, deep nesting and oversized arrays.

Does it send my response anywhere?

No, the analysis runs entirely in your browser.

How are formats detected?

By value patterns and key names: UUIDs, emails, URLs, ISO dates and date-times, Unix timestamps (seconds/milliseconds), IPs, JWTs, base64 blobs and id-like strings.

Is there a size limit?

Documents up to 2 MB are accepted; the analysis stops after 200 000 nodes and tells you if it did.

Technical notes

Paths with array indices are normalised to [*] so the same field across items is compared. Format detection combines value patterns with key names (e.g. a 10-digit number named createdAt is treated as a Unix timestamp).