MySQL JSON Path Tester

MySQL JSON paths look simple until one returns NULL, wraps a single value in an array or fails with ERROR 3143 Invalid JSON path expression. Paste a JSON document and one or more paths, one per line. For each valid path you get the result of JSON_EXTRACT, the unquoted ->> value and JSON_CONTAINS_PATH, and every location the path matched (for example $.items[1].sku), so you can see what * , ** and ranges such as [1 to last] really select. You also see whether the path can be used in JSON_SET, JSON_REPLACE and JSON_REMOVE or would fail with ERROR 3149. Invalid paths get the same ERROR 3143 message and character position MySQL reports, a marker under the problem and the likely fix, such as quoting a key that contains a space or a hyphen. Invalid documents get the ERROR 3141 message and position. Results use MySQL's own output format: keys sorted the way MySQL stores them, ", " separators and MySQL's number formatting. The behavior was checked against MySQL 8.4.11 with 20,000 generated paths and 3,048 number and invalid-document cases, with no differences. Everything runs in your browser; the document is never uploaded.

JSON document

JSON paths, one per line

Column name in generated SQL

Do more than mysql json path tester — meet Chat2DB

Chat2DB is an AI-powered SQL client for Windows, macOS and Linux. Write SQL in natural language, format and optimize queries automatically, and manage MySQL, PostgreSQL, Oracle and 20+ other databases in one workspace.

How to use

Paste the JSON document (a column value, an API payload or a sample row) into the first box.
Enter the paths to test, one per line, exactly as they appear inside the SQL string, and set the column name used in the generated SQL.
Click Test paths and read each result, the matched locations and any ERROR 3143 explanation, then copy the generated SELECT, JSON_TABLE or index statement.

Frequently asked questions

Why does MySQL say Invalid JSON path expression (ERROR 3143)?

The path text breaks MySQL's path grammar. The most common causes are a key with a space, hyphen or leading digit written without quotes ($.order-id must be $."order-id"), a negative index such as [-1] (use [last]), the JSONPath-style .. operator (MySQL uses $**.name), single quotes around a key, a range whose start is greater than its end ([3 to 1]) and a path that does not start with $. The position in the message counts bytes from the start of the path, so a multi-byte character before the problem shifts it. This tester shows the position with a marker and names the likely cause.

Why does JSON_EXTRACT return an array when I expect a single value?

A path that can match more than one value, because it contains *, ** or a range such as [0 to 2], always returns its matches wrapped in a JSON array, even when only one value matches: JSON_EXTRACT(doc, '$**.city') returns ["Lisbon"], not "Lisbon". Passing several paths to one JSON_EXTRACT call also wraps the results. If you need one row per match, use JSON_TABLE(doc, '$.items[*]' COLUMNS (...)). If you need a scalar, use an exact path such as $.items[0].sku.

What is the difference between ->, ->> and JSON_EXTRACT in MySQL?

col->'$.path' is shorthand for JSON_EXTRACT(col, '$.path') and returns JSON, so strings keep their double quotes. col->>'$.path' is JSON_UNQUOTE(JSON_EXTRACT(col, '$.path')) and returns plain text, which is what you usually want in WHERE clauses, comparisons and output. Both operators only accept a column on the left and a single path. When no value matches, all three return SQL NULL, which is not the same as a JSON null value. To try paths against real rows, connect to MySQL with Chat2DB (download it at chat2db.ai/download or open app.chat2db.ai) and run the generated queries.