You copied JSON from a log line, a config repo, or a chat thread. It might be minified, wrapped in single quotes, full of trailing commas, or sitting next to YAML that "should be equivalent." The JSON hub on Toolcore is not one mega-editor—it is a grid of small pages, each tuned to one job. This note is a plain map: start from your symptom, click through, and stop when the output looks right.
Everything linked below runs in your browser tab unless a card says otherwise. For a full card list, open the JSON tools hub.
Make messy JSON readable first
When the only problem is density—one long line from an API response—open the formatter, paste, and hit Format. If validation fails, read the error before you try anything fancier.
- JSON formatter— pretty-print, minify, validate, escape strings.
- JSON syntax rules— RFC 8259 reminders plus a strict parser check.
- JSON repair— when the paste is almost JSON (trailing commas, minor delimiter slips).
- JSON5 / JSONC → JSON— comments and relaxed keys from config files.
Compare two versions or explore structure
Line-based text diff lies about JSON—key order moves look like rewrites. Use structural compare when you care about fields, or a tree when you need to click through nested objects.
- Compare JSON— structural diff paths on two documents.
- JSON sort keys— stable key order before hashing or diffing again.
- JSON tree viewer— collapse branches, copy JSONPath-style paths.
- JSONPath & JMESPath— pull one field out of a large blob.
- jq filter— when you already think in jq pipelines.
Convert to or from other formats
Pick the page that matches the source format. Most converters keep a sample loaded so you can sanity-check direction (JSON → YAML vs YAML → JSON) before you paste production config.
- YAML ↔ JSON— the usual DevOps and docs handoff.
- JSON ↔ XML— legacy APIs and SOAP-ish payloads.
- CSV ↔ JSON— spreadsheet rows to objects (or back).
- TOML ↔ JSON— pyproject, Cargo, and similar configs.
- Excel (XLSX) → JSON and JSON → Excel— workbook import/export without a desktop app.
- NDJSON / JSON Lines— one value per line from streaming logs.
Validate against a contract
Syntax validation only asks "does it parse?" Schema tools ask "does it match the shape we promised?" Start from a sample payload, infer a draft schema, then tighten.
- JSON Schema validate— validate instances or infer draft-07 from data.
- JSON Schema → sample JSON— example objects from a schema file.
- JSON Pointer & Patch— RFC 6901 resolve and RFC 6902 diffs.
- JSON → OpenAPI draft— sketch an API description from example JSON.
Generate types or boilerplate from sample JSON
Codegen pages infer from one sample object (or the first row of an array). Treat output as a starting point—rename fields, nullability, and tags in your repo after export.
- JSON → TypeScript and JSON → Zod— front-end and runtime validation scaffolding.
- JSON → Go, Java, Python, Rust— pick the language card that matches your service.
Dozens more language and schema targets (Kotlin, C#, GraphQL SDL, Protocol Buffers, and others) live on the hub index— scroll the grid when you know the target ecosystem.
Clean up arrays and keys
When the JSON is valid but awkward—duplicate rows, null-heavy objects, or keys you want renamed in bulk—these pages save hand-editing.
- JSON cleaner (remove null) and rename keys
- group by field, unique by field, sort array by field
- flatten / unflatten— dot paths for nested objects.
When JSON is not the whole story
Base64 in a string field? Decode on Base64 & URL. Need a line diff on raw text instead of structure? Try text diff. Building a prefilled link for an assistant? See Save AI context with prefilled tool links. For Base64 or URL fields inside JSON, see Which encoding tool do I need?.
One habit worth keeping
Local processing means your paste does not need to round-trip through our servers for these transforms—but shared machines and screen shares still deserve redacted samples. Strip tokens, emails, and account ids before you demo a formatter in a meeting.