The first model was the GeForce 2 GTS, and the main difference between it and the GeForce256/DDR was the addition of a second texturing unit, thus more than doubling the texel fill rate. It was also clocked considerably higher than the previous generation GeForce cards.
Competitively, in OpenGL games (such as Quake III) it outperformed the Radeon and Voodoo 4/5 cards in all modes, but in DirectX games the Radeon was sometimes able to take the lead in 32-bit colour modes. This was attributed to the Radeon having slightly faster memory, and a better memory controller.
There were three more revisions on the GeForce 2 GTS core - the first was the GeForce 2 Ultra, launched in late 2000. This was basically a GeForce 2 GTS with much higher core and memory speeds, and put a definate lead between it and the Radeon and Voodoo 5. Some say it was intended to prevent 3dfx taking the lead with their Voodoo 5 6000 card, but in the event it wouldn't have been necessary, as the Voodoo 5 6000 was never launched.
The other two were the GeForce 2 Pro, and the GeForce 2 Ti. There were clocked at middling points between the GTS and Ultra cards, and intended to represent cheaper alternatives to the GeForce3, which never had a mainstream version.
However, Nvidia's real success story was with the GeForce 2 MX cards. These were cut down versions of the GeForce 2 GTS, with dual-monitor support (while there was a partial dual-monitor mode in the GTS, the monitors weren't independent, and would show the same image). It performed well enough to make it a viable mainstream alternative to the GTS (and its later revisions), and spawned two later versions - the MX200, which had a slower memory bus, and was designed for office work, and the MX400, which was a slightly faster MX. To this day, no graphics processor has matched to GeForce 2 MX (and its variants) in sales.