Nova 2 Lite vs Granite Embedding English r2

This comparison does not currently contain a protocol-matched coding benchmark for both Nova 2 Lite and Granite Embedding English r2, so Model Markets cannot name a coding leader from pricing, context size, or capability labels alone.

Only Nova 2 Lite has a directly sourced input price: $0.30 per million tokens. Only Nova 2 Lite has a directly sourced output price: $2.50 per million tokens.

Nova 2 Lite has the larger sourced context window. Nova 2 Lite supports 1,000K and Granite Embedding English r2 supports 8K.

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

Granite Embedding English r2 is the only model in this pair currently marked as self-hostable. Nova 2 Lite is not marked open weight; Granite Embedding English r2 is open weight.

Nova 2 Lite is documented with image input; Granite Embedding English r2 is not documented with image input. This reflects supported input modalities, not vision quality.

Neither has a larger sourced maximum output. Nova 2 Lite is 66K and Granite Embedding English r2 is —.

Nova 2 Lite: tool calling and image input. Granite Embedding English r2: none of these features are definitively sourced. Feature support does not establish relative quality.

Nova 2 Lite has 1 sourced provider route; Granite Embedding English r2 has 0, so Nova 2 Lite has broader tracked availability.

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.