Obfuscate JSON

Every JSON payload you hand to someone else carries a copy of your data model: the key names describe your domain, your naming conventions and your internal vocabulary. A vendor demo payload, a support screenshot, a bug report or a dataset sample all leak that shape — and once the field names are out, they are out. The fix is a reversible disguise: keep the structure, keep the types, keep the values, but replace every key name with something meaningless.

Obfuscate JSON does exactly that in one click. Keys become sequential names (a, b, c…) or random identifiers, the nesting, arrays and ordering are preserved exactly, and a mapping table records every rename for restoration. Keep specific keys readable for systems that need them, scramble string values when the content itself should be disguised, and re-roll the random names as often as you like. All client-side, nothing uploaded.

• Two naming styles: Deterministic sequential names (a, b, c…) or random base62 identifiers that defeat pattern-guessing.
• Structure preserved exactly: Nesting, arrays, types, ordering and values are untouched — only key names change.
• Keep-keys list: Named keys such as id or version pass through readable at every level.
• Scramble strings: Replace string values with placeholder text when the content itself must be disguised.
• Full mapping table: Every original key with its replacement, for mechanical restoration.
• Re-roll button: Regenerate random names at any time for a fresh disguise.
• Pretty or compact output: Match the output style to the consumer.
• Copy and download: Export the obfuscated document or the mapping table.
• Private: All processing happens in your browser; nothing is uploaded.

1. Paste your JSON into the input pane.
2. Pick a naming style — sequential for a clean mapping, random for a stronger disguise.
3. Add keep-keys (comma-separated) for fields that must stay readable, and toggle value scrambling if needed.
4. Generate — the obfuscated document appears alongside its mapping table.
5. Copy or download the document; keep the mapping table for restoration.

Example 1 — Vendor demo payload. A SaaS hands a fake-but-realistic payload to a partner for integration testing. Key renaming removes customer.payment.card_number from the visible schema; the partner integrates against a, b, c keys and learns nothing about the real data model.

Example 2 — Support ticket. A support engineer pastes a config dump into a ticket that will be shared with a third-party vendor. Obfuscated keys plus scrambled values keep the internal vocabulary and the content private while preserving the exact structure the vendor needs to diagnose.

Example 3 — Screenshot and presentation. An architecture review shows a real response on screen. Random-key obfuscation makes the fields unreadable at a glance while the audience still sees nesting depth, array sizes and value types.

Example 4 — Dataset publication. A team publishes a sanitized sample of an order export. Sequential key renaming plus the mapping table gives the research team a usable structure and the data team a one-line restore script.

• Share structure, not schema: Payloads circulate without leaking your field names or conventions.
• Reversible by design: The mapping table restores the original keys mechanically.
• Compatible output: Types, nesting and ordering survive, so tooling and tests keep working.
• Selective control: Keep-keys and value scrambling tailor the disguise to the audience.
• Instant and private: Client-side generation, live preview, no uploads.