Advertise here
#56

Nemotron 3 Super

NVIDIA Release: 2026-03-11 Tested on: 2026-04-20 17:48 nvidia/nemotron-3-super-120b-a12b::medium
120B total (12B active)MoEWeights available
(medium) (none)

Summary

Nemotron 3 Super scores 6.7 on AI BENCHY and ranks #56. It has N/A reliability, a 55.6% pass rate, $0.000 total cost, and 19.06s average response time.

What makes Nemotron 3 Super unique: Its total benchmark cost is unusually low for its score range.

Model facts

Researched on 2026-08-12

Reported
Parameters
120B total (12B active)
Architecture
MoE
Availability
Weights available
License
NVIDIA Open Model

Score

6.7

Consistency

8.7

Reliability

N/A

Total Output Tokens

41,715

Total Input Tokens

0

Input Price

$0.090 / 1M

Output Price

$0.450 / 1M

Tests Correct

Wrong Tests: 9

Attempt pass rate: 55.6%

Benchmark coverage: 52/54 attempts

Flaky tests

3

Flaky tests had mixed outcomes across runs (at least one pass and one fail).

Response Time (avg)

19.06s

Response Time (max): 87.80s

Response Time (total): 305.04s

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#56 Nemotron 3 Super

medium
Cost
$0.000
Time
272.6s
Tokens
5,296 tok

Run history

Tested on Score Reliability Tests Correct Total Cost Compare
2026-08-14 02:37 Re-test 5.7 9.1 $0.050 Compare
2026-07-16 23:04 New test added 5.7 8.7 $0.066 Compare
2026-06-04 14:07 New test added 5.8 10.0 $0.021 Compare
2026-05-22 00:27 Suite changed 5.9 10.0 $0.001 Compare
2026-05-08 15:31 Suite changed 6.1 10.0 $0.001 Compare
2026-04-20 17:48 First recorded run 6.7 N/A $0.000 Current run

Charts

Choose the first model, then click a second model to open a side-by-side page.

Total Output Tokens

Score vs Total Output Tokens

Quick Compare

Category Breakdown

Category Score Consistency Tests Correct
Anti-AI Tricks 10.0 10.0
Coding 3.0 10.0
Combined 10.0 10.0
Data parsing and extraction 10.0 10.0
Domain specific 2.9 4.4
General Intelligence 3.8 9.9
Instructions following 7.2 6.5
Puzzle Solving 3.5 9.8
Tool Calling 10.0 10.0

Compared models