Nvidia

Llama-3.1-Nemotron-Ultra-253B-v1

Name: Llama-3.1-Nemotron-Ultra-253B-v1
Brand: Nvidia

Canonical ID: nvidia/nvidia/llama-3.1-nemotron-ultra-253b-v1

More Nvidia models More llama-3.1-nemotron-ultra-253b-v1 family models

Flagship Nemotron model for high-throughput reasoning and complex agents

Reasoning: Yes Tool use: Yes Open weights

Overview

text

Provider: Nvidia
Family: llama-3.1-nemotron-ultra-253b-v1
Release date: 2024-07-01
Last updated: 2025-09-05
Input modalities: text
Output modalities: text

Pricing and Limits

Token pricing

Input: $0/1M
Output: $0/1M
Cached input: N/A

Context and output

Context window: 131,072
Max output tokens: 8,192
Qualified ID: nvidia/nvidia/llama-3.1-nemotron-ultra-253b-v1

Model Summary

Llama-3.1-Nemotron-Ultra-253B-v1 is a model listing in the Nvidia provider catalog. It is categorized under the llama-3.1-nemotron-ultra-253b-v1 family.

The model supports text modalities across input/output paths. Reported context capacity is 131,072. Pricing is listed at $0/1M input and $0/1M output.

Same Family: llama-3.1-nemotron-ultra-253b-v1

Comparable models in the same family across providers.

Browse family

Model	Provider	Context	Input	Output	Flags
NVIDIA: Llama 3.1 Nemotron Ultra 253B v1 kilo/nvidia/llama-3.1-nemotron-ultra-253b-v1	Kilo Gateway	131,072	$0.6/1M	$1.8/1M	Reasoning

Llama-3.1-Nemotron-Ultra-253B-v1

Overview

Pricing and Limits

Token pricing

Context and output

Model Summary

More From Nvidia

Active Speaker Detection

bevformer

BGE M3

ByteDance-Seed/Seed-OSS-36B-Instruct

Codegemma 1.1 7b

Codegemma 7b

Codellama 70b

Codestral 22b Instruct V0.1

Same Family: llama-3.1-nemotron-ultra-253b-v1