How to Fix JSON Parse Error — Step by Step (With Examples)

A JSON parse error stops your app cold. Whether you are seeing SyntaxError: Unexpected token in the browser console, a failed API response parse in Node.js, or a broken config file, the root cause is almost always one of a small set of syntax mistakes. This guide shows you exactly how to fix a JSON parse error step by step — with real examples of every common case.

These messages all mean the same thing: the JSON parser hit a character it did not expect according to the JSON specification. The position number tells you roughly where the error is, but finding the cause requires knowing what to look for.

The JSON specification is intentionally strict. Unlike JavaScript, it does not tolerate comments, trailing commas, single quotes, or any shorthand. Here are the six most common reasons why JSON is invalid:

{
  "name": "Alice",
  "age": 30,
  "role": "admin",
}

{
  "name": "Alice",
  "age": 30,
  "role": "admin"
}

{'user': 'Bob', 'active': true}

{"user": "Bob", "active": true}

A missing comma is one of the trickiest errors to spot by eye — especially in large JSON files. The parser error often points to the line after the missing comma, not the line where it belongs. Here is what it looks like:

{
  "firstName": "Jane"
  "lastName": "Smith",
  "email": "jane@example.com"
}

{
  "firstName": "Jane",
  "lastName": "Smith",
  "email": "jane@example.com"
}

{
  "tags": ["javascript" "typescript", "nodejs"]
}

{
  "tags": ["javascript", "typescript", "nodejs"]
}

The fastest way to debug a broken JSON file is to paste it into an online JSON error checker that highlights exactly where the problem is. Here is the exact process:

Here are five of the most frequently encountered JSON syntax errors with before-and-after examples for each:

{
  id: 1,
  name: "Product A",
  price: 19.99
}

{
  "id": 1,
  "name": "Product A",
  "price": 19.99
}

{
  // API version
  "version": "2.1",
  /* base URL */
  "endpoint": "https://api.example.com"
}

{
  "version": "2.1",
  "endpoint": "https://api.example.com"
}

{
  "count": NaN,
  "result": undefined,
  "ratio": Infinity
}

{
  "count": null,
  "result": null,
  "ratio": null
}

{"message": "He said "hello" to her"}

{"message": "He said \"hello\" to her"}

{
  "status": "ok",
  "data": {
    "id": 5
  }
}}

{
  "status": "ok",
  "data": {
    "id": 5
  }
}

AI tools like ChatGPT and Claude sometimes produce JSON with subtle errors — especially when asked to generate large JSON structures or when the output is truncated. Common AI-generated JSON problems include:

• Trailing commas on the last element (because models learn from JavaScript code more than strict JSON)
• Truncated output that ends mid-string or mid-array when the response hits a token limit
• Mixed quoting styles when the prompt included examples in different styles
• Missing closing brackets when nested structures are deep

Log aggregation tools like Splunk, Datadog, and CloudWatch also produce stringified JSON — where a JSON object is encoded as a string inside another JSON object. This creates double-escaped strings that need to be unescaped before they can be parsed.

// What you see in the log:
const raw = '{"level":"info","message":"{\"user\":\"alice\",\"action\":\"login\"}"}';

// Step 1: parse outer JSON
const outer = JSON.parse(raw);
// { level: 'info', message: '{"user":"alice","action":"login"}' }

// Step 2: parse the inner stringified JSON
const inner = JSON.parse(outer.message);
// { user: 'alice', action: 'login' }
console.log(inner.user); // 'alice'