Code, tree, grid, graph
Four synchronized ways to read the same document. Lazy tree expansion, virtualized rows and a structure map that stays responsive.
local-first · no account · no uploads
A JSON workbench for the work that comes after formatting.
Inspect structure, profile fields, find data-quality problems, query with real engines, compare versions, build transformations and debug API payloads. Everything runs on your device, in workers, without sending your documents anywhere.
- Editor
- Monaco, self-hosted
- Processing
- Web Workers
- Storage
- IndexedDB only
orders.json · profilelocal
Records
1,204
Max depth
6
Nullable
3 fields
Duplicate IDs
0
Missing email
12 (1%)
Quality score
86/100
$.orders[*].amountn = 1,204
Illustrative report layout. Real values come from the document you open — the studio never fabricates statistics.
One workspace, not a folder of unrelated converters
Views share one parsed document and one selection. A path found while profiling opens in the tree; a record filtered in the grid can be exported or transformed without re-importing anything.
Profiling & quality
Field cardinality, null frequency, missing fields, outliers and duplicate records — with custom rules, severity and a remediation note for every finding.
Schema intelligence
Infer JSON Schema from one or many documents, compare versions, detect breaking changes and validate with AJV Draft 7 / 2019-09 / 2020-12.
Diff, merge & patch
Structural comparison with identity matching, three-way merge with explicit conflict resolution, RFC 6902 and RFC 7396 export.
Queries & transformation
JSONPath, JMESPath, JSONata and a compact SQL dialect. Then a 21-step visual pipeline with per-step previews and generated code.
API & event streams
HAR and cURL import, snapshot drift detection, TypeScript types, JSONL timelines, correlation tracing and log-level filtering.
Also included:Privacy scanner & redactionSynthetic data generatorRelationship explorerDeveloper utilitiesOptional local Ollama assistant
Your documents stay on this device
There is no backend for document processing. Parsing, validation, profiling, queries, diffs and transformations execute in browser workers. Local persistence uses IndexedDB, and you can clear all application data from Settings at any time.
The public tool pages run the same local libraries as the studio. No document content is sent to an AI service. The optional assistant talks only to an Ollama endpoint you configure, and it is off by default.
- No mandatory account or cloud storage
- No analytics that inspect document content
- Strict CSP; no eval on user JSON or queries
- Installable PWA with an offline app shell
- Redaction mapping encrypted with AES-GCM when requested
- Exported files are created by your browser, not a server
Working tools, not landing pages
Each page runs a real local tool and links into the full studio for deeper work.
- Advanced JSON EditorA browser-based JSON editor with Monaco, duplicate-key detection, lossless large numbers, JSONC/JSON5 support, tree and grid views, and a full analysis workbench.
- Online JSON ViewerView, explore and understand JSON with a lazy tree, table view, structure graph and field statistics.
- JSON FormatterFormat and pretty-print JSON with 2, 4 or 8 space indentation, key sorting and precise error positions.
- JSON ValidatorValidate JSON syntax with exact line and column errors, duplicate-key detection and optional JSON Schema validation.
- JSON Diff ToolCompare two JSON documents structurally, ignore key order or selected paths, match arrays by id, and export RFC 6902 or RFC 7396 patches.
- JSON Schema ValidatorValidate JSON against Draft 7, 2019-09 or 2020-12 schemas with AJV.
- JSON to CSV ConverterConvert arrays of JSON objects to CSV with flattened nested columns and clear loss warnings.
- JSON Tree ViewerExplore JSON as a lazy, virtualized tree with type badges, array counts, paths and pinned nodes.
- Large JSON ViewerOpen large JSON files with worker parsing, virtualized views and progressive exploration.
- JSONPath TesterTest JSONPath expressions with filters, slices, unions and recursive descent.
- JSON Repair ToolRepair common JSON mistakes: comments, trailing commas, single quotes, unquoted keys and Python literals.
Limits, stated plainly
A tool that hides its boundaries is harder to trust. These are the current constraints; the documentation covers each in more detail.
- Standard JSON is parsed as one document; JSONL is parsed line by line. Streaming standard JSON is not implemented.
- The SQL engine supports SELECT, WHERE, GROUP BY, ORDER BY, LIMIT and basic aggregates. It does not support joins or subqueries.
- jq is not bundled. JSONata and JMESPath cover the maintained query-engine options.
- Schema validation uses the standard AJV vocabularies; custom vocabularies require code changes.
- Very large documents are bounded by browser memory. The app reports real sizes and timings rather than promising unlimited files.
Open a document and see for yourself
No sign-up. Drag a JSON file onto the studio, or start from a sample.