In short, no, but it's fairly close. GPS uses an elliptical earth model, so what you're looking for is the elevation with respect to the geoid. GPS uses the World Geodetic System (WGS) WGS84 model. Elevation is based on that geoid. Depending on location, this can be close, or not so close, compared with using mean sea level. See the link attached.
Albert Manfredi spelled out it nicely. That is why you cannot plot GPS heights on toposheets (Datum:Everest Spheroid in India). You can plot them on Google Earth
GPS is "trilateration" as opposed to "triangulation". Triangulation makes measured based on three points and gives you a location on a plane. (Triangulation assumes you are trying to determine the location of a point on a flat surface. Trilateration requires four points because you need elevation as well.)
Simply, if you know your distance from a single point, (I am 100 feet from the Washington Monument), then on a flat surface I can be anywhere on a circle that has a radius of 100 feet. But if you care about 3 dimensions then the fact is that you are somewhere on a sphere that has a radius of 100 feet. I could be underground 100 feet!
The reality is that the elevation component of GPS is fairly insignificant compared to the other components. GPS also relies on a theoretical model of the earth as an ellipse (as noted above) so Mr. Manfredi's answer is most accurate.
The raw elevation measured by the GPS is the height above the WGS84 ellipsoid. This elevation is different from the orthometric height measured based on the geoid (global model of the gravity field fitted to the heigth of the mean sea level) or based on a levelling network (that uses the mean sea level at one point). Sometimes the GPS devise will automatically compute an orthometric height, but if you are in HAE the difference can be quite large (-100 in Southern India, + 80 in New Guinea)
The location provided by GPS is based not on "triangulation" which is a point on a plane but "trilateration" . The answer is a point in three-dimensional space. When you "triangulate" the answer is a point on a plane.
That said, the elevation is not the actual altitude about sea level or above the actual land, but the distance off of a perfect ellipse that is the estimate of the size of the earth.