Skip to content
aiqbench

Nvidia · model passport

Llama 3 1 Nemotron Ultra 253b V1

llama-3-1-nemotron-ultra-253b-v1
  • Open weights
  • Nvidia Open Model licence
  • 253.4B params
  • Released 2025-04-08(curated)
8/27axes covered
Overall · LMArena
1,348#204 of 396
Best percentile
81.3MMLU-Pro (Artificial Analysis)
Cohort standing
49.1median 47.8

What the data says

Benchmarked on 8 of 27 tracked axes. Every figure below is the model's own published result on that board.

Weakest axes

  • 2.4 pct — BFCL v4 (function calling) (10.0%)

Uneven: 78.9 percentile points between its best and worst axis — 81.3 pct on MMLU-Pro (Artificial Analysis) versus 2.4 pct on BFCL v4 (function calling).

  • MMLU-Pro (Artificial Analysis)81.3 pct82.5%Verify →
  • LiveCodeBench (Artificial Analysis)67.9 pct64.1%Verify →
  • Instruction following52.8 pct1,351Verify →
  • Hard prompts52 pct1,375Verify →
  • Overall48.7 pct1,348Verify →
  • Coding48.6 pct1,391Verify →
  • IFBench (Artificial Analysis)36.1 pct38.2%Verify →
  • BFCL v4 (function calling)Dormant board2.4 pct10.0%Verify →
  • WebDev ArenaNot benchmarked
  • Vision ArenaNot benchmarked
  • Factuality ArenaNot benchmarked
  • Search ArenaNot benchmarked
  • Document ArenaNot benchmarked
  • Agent ArenaNot benchmarked
  • GPQA DiamondNot benchmarked
  • AIME (math)Not benchmarked
  • FrontierMathNot benchmarked
  • SimpleQA VerifiedNot benchmarked
  • ARC-AGI-2Not benchmarked
  • Humanity's Last ExamNot benchmarked
  • SimpleBenchNot benchmarked
  • LiveBenchNot benchmarked
  • SWE-bench Verified (live board)Not benchmarked
  • Terminal-Bench 4.0Not benchmarked
  • HHEM Hallucination (Vectara)Not benchmarked
  • SWE-bench VerifiedNot benchmarked
  • Terminal-BenchNot benchmarked

Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.

How to run it

Run it yourself

Repositorynvidia/Llama-3_1-Nemotron-Ultra-253B-v1
Parameters253.4B
ArchitectureDeciLMForCausalLM
LicenceNvidia Open Model

Transformers

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "nvidia/Llama-3_1-Nemotron-Ultra-253B-v1"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

Ollama

We don't sync Ollama tags, so we won't print a command we can't verify — a wrong one would pull the wrong model. Search the Ollama library →

Unsloth fine-tuning recipes are model-specific and we don't derive them here. Unsloth docs →

Identifiers, limits and prices are synced from the upstream source shown above. Any field we haven't synced is left out entirely rather than guessed.

  • llama-3.1-nemotron-ultra-253b-v1 stands at 49.1 (raw mean 48.7) — measured on 8 of 27 axes.

    standingtemplateVerify →

  • llama-3.1-nemotron-ultra-253b-v1 stands at 49.1 versus a cohort median of 47.8 across 151 models released within 180 days — above the peer median.

    cohorttemplateVerify →

Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →