Skip to content
Trending storyDeveloping

Artificial Analysis measures Nemotron 3.5 Super VL at ~630 tokens/s at 100k-token prompts

1 article1 sourcesince Oct 15Last article 1h ago ·

Overview

AISummary of 1 article

Artificial Analysis measured Nemotron 3.5 Super VL at about 630 output tokens/s with 100k-token prompts, the fastest of any model scoring 20 or more on its Intelligence Index at that length.

The test ran on a pre-release DeepInfra endpoint serving NVFP4 weights.

The firm reports the model is about 5x Inkling Small (124 tokens/s), 2.7x Nemotron 3 Ultra (235 tokens/s), and 4.5x Nemotron 3 Super (140 tokens/s). It also returns its first token in 1.8 seconds, against 4.4 seconds for Nemotron 3 Ultra and 8.5 seconds for Inkling Small.

Written by AI from the articles below · updated Oct 15, 12:02 PM ET

Check the sources:

Latest developmentOct 15, 11:51 AM ET
Nemotron 3.5 Super VL reaches 630 tokens/s at 100k-token prompts

Article timeline

The articles in this story. Times are ET.

Oct 15
  1. Artificial AnalysisOfficial
    Nemotron 3.5 Super VL reaches 630 tokens/s at 100k-token prompts

    AIArtificial Analysis measured Nemotron 3.5 Super VL at about 630 output tokens/s with 100k-token prompts on a pre-release DeepInfra endpoint serving NVFP4 weights, the fastest of any model scoring 20 or more on the Intelligence Index at that length. It is about 5x Inkling Small (124 tokens/s), 2.7x Nemotron 3 Ultra (235), and 4.5x Nemotron 3 Super (140), and returns its first token in 1.8s versus 4.4s and 8.5s for Nemotron 3 Ultra and Inkling Small.

    Image from @ArtificialAnlys's post

Heat trend

Not enough continuous observations to show a trend yet.