In a very big molecule [1] that doesn't move, you have so many parts that you can have a reasonable good definition of the temperature.
But in a very small molecule/atom, let's pick a single helium atom, you don't have a good definition of temperature. You can define the temperature of a gas of Helium using the average kinetic energy of the atoms, but you must consider that the whole gas is not moving.
If you put a balloon with Helium in a car, and the car moves at 100mph, the Helium is not hotter, it is moving. So if you can only see a single atom of Helium, you can't be sure if it's moving because it's hot or all the gas is moving. So you don't have a good definition of temperature. [2]
For a molecule with a few atoms (let's say 5 or 10) it's more difficult to be sure if there is a good definition of temperature or not, so I prefer to ignore the intermediate case.
[1] My first idea of a big molecules was DNA, because it's big and well known. But DNA is usually surrounded by water, lots of water, and ions and auxiliary proteins, and a lot of stuff. It's very difficult to isolate a true molecule of DNA alone.
An easy example of a big molecule is bakelite https://en.wikipedia.org/wiki/Bakelite that is the plastic of the old phones. Your old phone case was a gigantic single molecule, and the tube case was another. So they clearly had a temperature.
[2] At low temperatures you can assume that a Helium atom is a ball without internal structure. If you increase the temperature of the gas enough (200000K?), the electrons in each Helium atom start to jump between levels and you have some interesting internal structure and may try to define a temperature. But it's still too few parts and too short lived to make me comfortable to define the temperature of a single atom.