OpenRating is an independent, open-source AI rating platform. Join the community

OOpenRating
All models

NVIDIA · Released Mar 11, 2026

Nemotron 3 Super 120B A12B (Reasoning)

Nvidia's reasoning-optimized model featuring verified performance on the OpenRating benchmark index, balanced for high throughput and cost efficiency.

Open weights

OpenRating Index

0.0

0–100 composite

Reasoning Index

0.0

0–100 composite

Cost / Task

$0.00

benchmark run

Output speed

131 tok/s

blended tok/s

Time to first token

32.6s

median latency

Input price

$0.3

per 1M tokens

Output price

$0.9

per 1M tokens

Context window

1M

tokens

Strengths

  • Intelligence Index score of 0.0 on independent evaluations

  • 131.0 tokens/sec median generation throughput with sub-second response times

  • 1000k token context window supporting complex multi-document reasoning

How we rate Nemotron 3 Super 120B A12B (Reasoning)

OpenRating scores every model on the same axes. The OpenRating Index is a weighted composite of reasoning, knowledge, and agentic benchmarks, normalised to a 0–100 scale. Speed is blended output measured in tokens per second under standardised load, with time-to-first-token reported separately. Prices are list input and output prices per million tokens, and Task Cost measures end-to-end workload spending.

OpenRating Index

0.0

Reasoning Index

0.0

Token Pricing

Input per 1M tokens

$0.3

Output per 1M tokens

$0.9

Cost of 1M input + 1M output

$1.2

Task Cost Economics

Empirical cost breakdown for a standard OpenRating Index benchmark task ($0.00 total).

Reasoning (Chain of Thought)
$0.00
Answer generation
$0.00
Cache write overhead
$0.00
Cache hit queries
$0.00
Input payload
$0.00