readable AI benchmarks (simplified)

readable AI benchmarks

NVIDIA

Nemotron 3 Super 120B A12B (Reasoning)

reasoning · open weights · Mar 11, 2026

Nemotron 3 Super 120B A12B (Reasoning) has a Quality Score of 24.5, ranking 335th among 571 scored models. Per 1M tokens, pricing is $0.300 input, $0.900 output, and $0.150 cached input. It accepts text input, outputs text, and has a 1M token context window. Its Reliability Score is 29.3, ranking 329th among 573 scored models, below average (total average: 36.5). Its Value Score is 20.1, low compared with the total Value average of 31.4.

Quality Score24.5coverage 100.0%: AA Intelligence and AA-Omniscience outcomes available
Value Score20.1
Factual reliability29.3
Cache Discount50%

Model specification

Reasoning
reasoning
Input modalities
text
Output modalities
text
Context window
1M tokens
Weights
open weights

Token prices USD per 1M tokens

Input
$0.300
Output
$0.900
Cached Input
$0.150
Cache write
not available
Cost per task
$1.64

Capability

Intelligence
12.8
Coding
not available
Agentic
not available
Omniscience
-41.5
Correct
24.3%
Blended price
$0.255

Answer outcomes

Fully graded outcomes (Correct + Incorrect): 90.1%

Correct 24.3%Incorrect 65.8%Partial / not attempted 9.9%

Artificial Analysis benchmarks

GDPval-AA v2
0%
τ³-Banking
10%
SciCode
36%
Humanity’s Last Exam
21%
GPQA Diamond
80%
CritPt
3%
AA-Omniscience
29%
AA-LCR
66%

Similar models

Support me! Patreon