LIVE Cheapest: GLM-4.7-Flash $0.000/Mtok in 153 models tracked Updated Invalid Date
Invalid Date
▍LLMPrice$/Mtok
⌕
Pricing / NVIDIA / Llama Nemotron Ultra 253B

Llama Nemotron Ultra 253B

by NVIDIA · 253B parameters

Current open weights Open weights mid tier

Pricing · per 1M tokens

Input
$0.600
per million tokens
Output
$3.60
per million tokens
Blended avg cost*$2.10/Mtok

Overview

Large NVIDIA-tuned Llama model. $0.60/$3.60 per 1M on hosted platforms.

Specifications

ProviderNVIDIA
Context window128K tokens
Modalitytext
Parameters253B
Open sourceYes — open weights available
ReleasedJan 1, 2025
StatusCurrent
Last updatedJun 24, 2026
Tagsopen-weights reasoning
Official pricing page →

Cost calculator

1K tokens (in)$0.0006
1K tokens (out)$0.0036
100K tokens (in)$0.06
100K tokens (out)$0.36
1M tokens (in)$0.600
1M tokens (out)$3.60
10M tokens (blended)$21
Full calculator →

At a glance

Input$0.600/M
Output$3.60/M
Blended$2.10/M
Context128K
Tiermid

Related models

Llama 3.1 8B
Meta · $0.065/M blended
Pixtral 12B
Mistral · $0.1/M blended
Nemotron 70B Instruct
NVIDIA · $0.1/M blended
Llama 4 Scout
Meta · $0.225/M blended
Mixtral 8x7B Instruct
Mistral · $0.54/M blended
Llama 3.3 70B
Meta · $0.69/M blended