JSON Minimal Mode

JSON Minimal Mode reduces a large JSON document to only the fields your current task requires. Rather than reformatting whitespace like a minifier, it restructurally filters: you include the paths you want or exclude the paths you do not, and the result is a smaller, task-focused payload. When you are sending JSON to an LLM, sharing a subset with a third party, or pulling metrics from a deeply nested response, minimal mode removes everything irrelevant without touching the data you need.

1. Paste your full JSON document into the input panel.
2. Choose whether to work with an include list (keep only these paths) or an exclude list (remove these paths).
3. Enter the dot-notation paths for the fields you want to keep or remove — for example, user.id, user.email, meta.
4. Review the filtered output to confirm the right fields survived.
5. Copy or download the minimal payload for use in your prompt, export, or downstream tool.

• Token waste from sending entire API responses to LLMs when only two or three fields are relevant to the prompt
• Privacy exposure when sharing JSON samples that contain PII alongside the fields you actually want to show
• Payload bloat in analytics pipelines where downstream consumers only need a subset of event properties
• Cognitive overload when reviewing or debugging a response that contains dozens of fields unrelated to the current task
• Repetitive manual extraction of the same fields from multiple payloads

A typical API user response contains far more than most tasks need:

{
  "user": {
    "id": "u-7741",
    "name": "Sonia Park",
    "email": "[email protected]",
    "passwordHash": "...",
    "createdAt": "2023-11-14T08:22:00Z",
    "lastLogin": "2026-04-20T14:05:33Z",
    "preferences": {
      "theme": "dark",
      "notifications": true
    },
    "internalFlags": {
      "betaAccess": false,
      "reviewQueue": true
    }
  }
}

Including only user.id, user.name, and user.email produces a three-field object ready for a prompt or a public-facing log without exposing the hash or internal flags.

• LLM prompt preparation. You have a 4 KB API response but your prompt only needs the id, status, and items fields. Filter to those paths before inserting into context to conserve tokens and improve model focus.
• Privacy-safe sharing. A teammate needs a real payload for debugging but it contains PII. Exclude email, phone, and address fields before sharing.
• Analytics handoffs. Your event log includes dozens of instrumentation fields. An analyst only needs eventType, timestamp, and userId. Produce that slice without rewriting the original emitter.
• Cleaner documentation examples. Strip internal and autogenerated fields from a real payload so the example in your docs only shows what external consumers actually see.

• JSON Trimmer — use first if the payload contains comments or trailing commas that would prevent clean parsing
• JSON Formatter — reformat the filtered output for readability if you are using it for documentation or review
• Pretty vs Minified JSON — understand the difference between structural reduction and whitespace compression