Artificial Analysis benchmarks NVIDIA Nemotron Nano 12B v2 VL
Artificial Analysis has published a benchmark page for NVIDIA's Nemotron Nano 12B v2 VL, a two-model release, but the public snapshot reports a latency figure and no capability scores.
Artificial Analysis has published a release page for NVIDIA Nemotron Nano 12B v2 VL. The page lists two models under the release, each with its own intelligence, performance and pricing profile.
The only figure stated in the text is latency. The non-reasoning variant of Nemotron Nano 12B v2 VL has a time to first answer token of 1.49 seconds, the lower of the two models in the release. The page does not give the second model's latency.
The page explains how Artificial Analysis compares models. Non-reasoning models are ranked only against other non-reasoning models. Reasoning models are ranked against both kinds. Open-weights models are grouped by parameter count, with "Small" covering 4B to 40B parameters. Proprietary models are grouped by blended price, using a 3:1 input-to-output ratio.
The page lists the evaluations behind its capability indexes, including AA-Omniscience v1.1, AA-Briefcase v1.1, HLE-Diamond, AutomationBench-AA, GDP.pdf, MLCR-AA, CritPt v2 (Hard Subset), Terminal-Bench 4.0, Terminal-Bench-Science 0.1 and FrontierCode-AA 2.0.
The text includes no scores for the model on these indexes and no figures for price, cost per task or output speed. It also does not say whether the weights are open, how the two models differ beyond the non-reasoning label, or which vision or multimodal tasks the "VL" model was tested on.