Paste a JSON array of objects, get a PostgreSQL DDL with auto-detected types.
Open the Schema Inferrer →When your data lives in JSON — API responses, log lines, NoSQL exports — the schema inferrer flattens the top-level keys into columns and infers PostgreSQL types from the values. Numbers become INT/BIGINT/NUMERIC, true/false becomes BOOLEAN, ISO 8601 strings become DATE or TIMESTAMP, UUID-shaped strings become UUID.
Sample values and the dialect-native type they map to.
| Sample value | Inferred type |
|---|---|
| "value": 42 | INT |
| "value": 9999999999 | BIGINT |
| "value": 12.45 | NUMERIC(10,2) |
| "value": true | BOOLEAN |
| "value": "2024-03-15" | DATE |
| "value": "2024-03-15T09:30:00Z" | TIMESTAMP |
| "value": "a3bb189e-…" | UUID |
| "value": null | (nullable) |
[
{"id": 1, "name": "Alice", "score": 9.4, "active": true},
{"id": 2, "name": "Bob", "score": 7.2, "active": false}
]
CREATE TABLE users (
id INT NOT NULL,
name VARCHAR(16) NOT NULL,
score NUMERIC(2,1) NOT NULL,
active BOOLEAN NOT NULL
);
No signup, no install — runs entirely in your browser.
Open the Schema Inferrer →