readable AI benchmarks (simplified)

readable AI benchmarks

NVIDIA

Nemotron 3 Ultra 550B A55B (Reasoning)

reasoning · open weights · Jun 4, 2026

Nemotron 3 Ultra 550B A55B (Reasoning) has a Quality Score of 42.5, ranking 166th among 571 scored models. Per 1M tokens, pricing is $0.600 input, $2.50 output, and $0.160 cached input. It accepts text input, outputs text, and has a 262,144 token context window. Its Reliability Score is 49.8, ranking 131st among 573 scored models, above average (total average: 36.5). Its Value Score is 32.5, around average compared with the total Value average of 31.4.

Quality Score42.5coverage 100.0%: AA Intelligence and AA-Omniscience outcomes available
Value Score32.5
Factual reliability49.8
Cache Discount73%

Model specification

Reasoning
reasoning
Input modalities
text
Output modalities
text
Context window
262,144 tokens
Weights
open weights

Token prices USD per 1M tokens

Input
$0.600
Output
$2.50
Cached Input
$0.160
Cache write
not available
Cost per task
$0.598

Capability

Intelligence
22.9
Coding
not available
Agentic
not available
Omniscience
-0.4
Correct
22.6%
Blended price
$0.482

Answer outcomes

Fully graded outcomes (Correct + Incorrect): 45.6%

Correct 22.6%Incorrect 23.0%Partial / not attempted 54.4%

Artificial Analysis benchmarks

GDPval-AA v2
26%
τ³-Banking
14%
SciCode
40%
Humanity’s Last Exam
28%
GPQA Diamond
87%
CritPt
3%
AA-Omniscience
50%
AA-LCR
79%

Similar models

Support me! Patreon