modelbenchmark.io

nemotron-3-ultra-nvfp4

NVIDIA · nemotron-3-ultra-nvfp4

Compare

Nemotron-3-Ultra-550B-A55B-NVFP4 is a frontier-scale large language model (LLM) trained by NVIDIA, designed to deliver strong agentic, reasoning, and conversational capabilities. It is optimized for the most demanding workloads, including complex multi-step agents, long-context analysis, and high-accuracy reasoning over code, math, and science. The model employs a hybrid Latent Mixture-of-Experts (LatentMoE) architecture, utilizing interleaved Mamba-2 and MoE layers, along with select Attention layers. Like the Super model, the Ultra model incorporates Multi-Token Prediction (MTP) layers for faster text generation and improved quality, and it is trained using an NVFP4 pre-training recipe to maximize compute efficiency. The model has 55B active parameters and 550B parameters in total.

Specification

most-agreed values

Context
262K
Max output
262K
Released
2026-06-13
Knowledge cutoff
Retires
Open weights
no
Input
text
Output
text

Price

US dollars per million tokens · most-agreed

Input
$0.60
Output
$2.40
Cache read
$0.12

Available from 3 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
fireworks_ai$0.60$2.40$0.12262K33K
fireworks-ai · accounts$0.60$2.40$0.119262K128K
requesty$0.60$2.40$0.12262K262K

More from NVIDIA

most-hosted first