Granite Embedding English r2 vs NVIDIA Nemotron 3 Super 120B A12B Base BF16

This comparison does not currently contain a protocol-matched coding benchmark for both Granite Embedding English r2 and NVIDIA Nemotron 3 Super 120B A12B Base BF16, so Model Markets cannot name a coding leader from pricing, context size, or capability labels alone.

Neither model has a directly sourced input price in this comparison. Neither model has a directly sourced output price in this comparison.

NVIDIA Nemotron 3 Super 120B A12B Base BF16 has the larger sourced context window. Granite Embedding English r2 supports 8K and NVIDIA Nemotron 3 Super 120B A12B Base BF16 supports 1,049K.

There is no overall benchmark winner: At least two independently verified, protocol-matched benchmarks are required for an overall winner.

Both models have the same recorded self-hosting status: supported. Granite Embedding English r2 is open weight; NVIDIA Nemotron 3 Super 120B A12B Base BF16 is open weight.

Granite Embedding English r2 is not documented with image input; NVIDIA Nemotron 3 Super 120B A12B Base BF16 is not documented with image input. This reflects supported input modalities, not vision quality.

Neither has a larger sourced maximum output. Granite Embedding English r2 is — and NVIDIA Nemotron 3 Super 120B A12B Base BF16 is —.

Granite Embedding English r2: none of these features are definitively sourced. NVIDIA Nemotron 3 Super 120B A12B Base BF16: none of these features are definitively sourced. Feature support does not establish relative quality.

Granite Embedding English r2 has 0 sourced provider routes; NVIDIA Nemotron 3 Super 120B A12B Base BF16 has 0, a tie.

There is no universal value winner. Compare the input and output prices above with the matched benchmark result for your workload: cheaper tokens can be offset by different quality, token usage, latency, or provider availability.