Free tool
JSON viewer.
Paste an API response or open a .json file and browse it as a tree. Fold what you don't need, search for the key you do, and click any node to copy the exact path to it.
- {6}
- service: "checkout"
- deploy: {4}
- sha: "9f3c1ab"
- env: "production"
- at: "2026-10-15T09:14:00Z"
- author: {2}
- checks: [3]
- 0: {3}
- 1: {3}
- 2: {4}
- headers: {2}
- content-type: "application/json"
- x-request-id: "req_8c1f"
- flags: {2}
- new_cart: true
- legacy_tax: false
- rollback: null
Click any row to get its path.
Publishes the tree to Comma as a page that folds and unfolds without any script. Readers comment on the exact branch, and every re-publish is kept as a revision. Free, no card.
Search that opens the tree
In a big response the field you want is three levels down inside the forty-second item of an array. Search here matches keys and values, then opens exactly the branches that lead to a hit and leaves everything else folded — so the answer is on screen without scrolling past the thirty-nine items that did not match.
Copy the path, not the value
Most of the time you open a JSON viewer to write code against the payload. Click a node and you get both spellings of where it lives: the JSONPath ($.checks[2].error) for tools and test assertions, and the accessor (data.checks[2].error) that pastes straight into JavaScript or a jq filter. Keys like content-type are bracket-quoted for you. More on the syntax in the JSONPath glossary entry.
Viewer, formatter or table?
Use the viewer to navigate. If the file does not parse, the JSON formatter tells you the line, column and likely cause. If the payload is an array of records and the reader is a person rather than a programmer, JSON to HTML table or JSON to chart will say more than any tree. Whichever you pick ends as a Comma link, and an agent can publish the same thing at the end of a run over the MCP server.
Questions
- How do I view a JSON file as a tree?
- Paste the JSON or open a .json file. It is parsed in your browser and shown as a collapsible tree: objects and arrays fold and unfold, each shows how many children it has, and the first two levels open by default. Expand all and Collapse all work on the whole document.
- How do I find a key in a large JSON file?
- Type in the search box. Every key or value containing the text is highlighted and the tree opens exactly as far as needed to show each match — everything else stays folded, so a hit three levels down inside one array item is visible without unfolding the rest.
- How do I get the path to a value?
- Click any row. The bar under the tree shows its JSONPath ($.checks[2].error) and the same path as a JavaScript or jq accessor (data.checks[2].error), each with a copy button. Keys that are not valid identifiers are bracket-quoted, so content-type comes out as ["content-type"].
- What is the difference between a JSON viewer and a JSON formatter?
- A formatter re-indents the text so you can read it top to bottom. A viewer lets you navigate it — fold what you do not care about, search, and copy the path to the one field you need. For a parse error with a line and column, use the JSON formatter; for a 5,000-line API response, use the viewer.
- Does it read NDJSON / JSON Lines?
- Yes. If the input is not one JSON document but every line parses on its own, it is shown as an array with one entry per line.
- Can I share the tree with someone?
- Get a shareable link publishes the tree as a Comma report: the same folding tree, built from plain HTML details elements so it works without any script, with comments anchored to the branch someone is asking about. Free, no card.
- Is my JSON uploaded anywhere?
- No. Parsing, searching and rendering happen in this tab. The data only leaves your machine if you choose Get a shareable link, which publishes it as a Comma report whose access you control.
Related: JSON formatter · JSON to HTML table · JSON to CSV · YAML to JSON · XML to JSON · JWT decoder · NDJSON · all free tools