JSON Formatter & Validator

JSON Reference

What JSON is, what it isn't, and the day-to-day gotchas — number precision, single-quote rejection, the JSON-vs-JSON5-vs-JSONC distinction, and parsing in different languages.

The JSON data types

JSON vs JSON5 vs JSONC vs JSONL

This tool parses strict JSON. For JSONC, strip // and /* */ comments before pasting; for JSON5, also remove trailing commas.

Parsing JSON in different languages

const obj = JSON.parse(text)
const str = JSON.stringify(obj, null, 2)

// Type-safe with zod / valibot:
import { z } from 'zod'
const schema = z.object({ name: z.string() })
const parsed = schema.parse(JSON.parse(text))

import json

obj = json.loads(text)
text = json.dumps(obj, indent=2)

# With Pydantic v2:
from pydantic import BaseModel
class User(BaseModel):
    name: str
user = User.model_validate_json(text)

import "encoding/json"

type User struct {
    Name string `json:"name"`
}

var u User
err := json.Unmarshal(data, &u)
out, err := json.MarshalIndent(u, "", "  ")

use serde::{Deserialize, Serialize};

#[derive(Deserialize, Serialize)]
struct User { name: String }

let u: User = serde_json::from_str(&text)?;
let out = serde_json::to_string_pretty(&u)?;

$obj = json_decode($text, true);
$text = json_encode($obj, JSON_PRETTY_PRINT);

// With JSON_THROW_ON_ERROR (PHP 7.3+)
$obj = json_decode($text, true, 512,
    JSON_THROW_ON_ERROR);

# Pretty-print
echo "$text" | jq .

# Minify
echo "$text" | jq -c .

# Extract a field
echo "$text" | jq -r '.user.name'

Common gotchas