Size of json object? (in KBs/MBs)

An answer to the actual question should include the bytes spent on the headers and should include taking gzip compression into account, but I will ignore those things.

You have a few options. They all output the same answer when run:

If Using a Browser or Node (Not IE)

const size = new TextEncoder().encode(JSON.stringify(obj)).length
const kiloBytes = size / 1024;
const megaBytes = kiloBytes / 1024;

If you need it to work on IE, you can use a pollyfill

If Using Node

const size = Buffer.byteLength(JSON.stringify(obj))

(which is the same as Buffer.byteLength(JSON.stringify(obj), "utf8")).

Shortcut That Works in IE, Modern Browsers, and Node

const size = encodeURI(JSON.stringify(obj)).split(/%..|./).length - 1;

That last solution will work in almost every case, but that last solution will throw a URIError: URI malformed exception if you feed it input containing a string that should not exist, like let obj = { partOfAnEmoji: "👍🏽"[1] }. The other two solutions I provided will not have that weakness.

(Credits: Credit for the first solution goes here. Credit for the second solution goes to the utf8-byte-length package (which is good, you could use that instead). Most of the credit for that last solution goes to here, but I simplified it a bit. I found the test suite of the utf8-byte-length package super helpful, when researching this.)

For ascii, you can count the characters if you do...

JSON.stringify({
  'a': 1,
  'b': 2
}).length

If you have special characters too, you can pass through a function for calculating length of UTF-8 characters...

function lengthInUtf8Bytes(str) {
  // Matches only the 10.. bytes that are non-initial characters in a multi-byte sequence.
  var m = encodeURIComponent(str).match(/%[89ABab]/g);
  return str.length + (m ? m.length : 0);
}

Should be accurate...

var myJson = JSON.stringify({
  'a': 1,
  'b': 2,
  'c': 'Máybë itß nºt that sîmple, though.'
})

// simply measuring character length of string is not enough...
console.log("Inaccurate for non ascii chars: "+myJson.length)

// pass it through UTF-8 length function...
console.log("Accurate for non ascii chars: "+ lengthInUtf8Bytes(myJson))

/* Should echo...

Inaccurate for non ascii chars: 54
Accurate for non ascii chars: 59

*/

Working demo