NVIDIA

Llama 3.1 Nemotron Ultra 253B

CTX128K

44Quality

Reasoning · Tools

InputFree

OutputFree

Flagship Nemotron model for high-throughput reasoning and complex agents

FreeNVIDIA Open Model license534/1044

Drag to turn

NVIDIATextCommon

Llama 3.1 Nemotron Ultra 253B

Flagship Nemotron model for high-throughput reasoning and complex agents

nvidia/llama-3.1-nemotron-ultra-253b-v1

Type: Text
Quality: 44 · Top 56% on the LMArena Text
Context window: 128,000 tokens
Max output: 16,384 tokens
Input price: Free
Output price: Free
Cache read: —
Cache write: —
Parameters: 253B (253,401,268,224)
Downloads: 1.6K in the last 30 days
Likes: 360
Access to weights: Open download
Released: Apr 7, 2025
Knowledge cutoff: —
Takes in: Text
Puts out: Text
Reasoning: Yes
Tool calling: Yes
Structured output: No
File attachments: No
Access: Free
License: NVIDIA Open Model license
Rarity: Common
Set number: 534/1044

Open NVIDIA docs Weights on Hugging Face Quality 44 Where to run it 1 Meet its creature Battle North Small Translate

No matching Hugging Face repo was found, so this shows NVIDIA’s usual open-weight license. Check the model card before relying on it. Specs from models.dev, downloads from Hugging Face, last changed Oct 19, 2026. Official source: docs.api.nvidia.com.

How good is it?

Public leaderboard scores for Llama 3.1 Nemotron Ultra 253B, refreshed daily from Epoch AI and LMArena.

Top 56% of the 416 models on the LMArena Text: #235.

Quality is the share of a leaderboard’s models this one beats, from 0 to 100. Chat models are rated on Epoch AI’s Capabilities Index when it lists them, otherwise on LMArena; image and video models on LMArena’s image and video arenas.

Leaderboards

LMArena Text#235 of 416
People compare two anonymous answers and vote for the better one.1,320 rating
LMArena Text · Coding#249 of 411
The text arena’s coding prompts only.1,312 rating
LMArena Text · Math#203 of 400
The text arena’s math prompts only.1,360 rating
LMArena Text · Creative writing#210 of 414
The text arena’s creative writing prompts only.1,314 rating

Scores from LMArena, licensed CC BY 4.0. Boards list models under their own names, so a close match is shown with the name it was scored under. Credits

Where to run it

1 provider sells Llama 3.1 Nemotron Ultra 253B through an API. Prices are per million tokens, cheapest first.

Free to try1 providerNvidia
From the labNvidiaFree in · Free out

API providers

Provider	Input	Output	Context
NvidiaOfficialFree`nvidia/llama-3.1-nemotron-ultra-253b-v1`	Free	Free	128K

From models.dev and Hugging Face, checked daily. Resellers’ listings can lag behind their price pages, so confirm with the provider before you commit.

Face off

Current text models from other labs at a similar output price.