BazaarLinkBazaarLink
Sign in

Llama 3.1 Nemotron Ultra 253b V1 — Retired model

Retired modelThis model has reached end of life and is no longer available through BazaarLink. Historical specifications and independent benchmarks remain below. Retired on May 1, 2026.

Llama-3.1-Nemotron-Ultra-253B-v1 is a large language model (LLM) optimized for advanced reasoning, human-interactive chat, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta’s Llama-3.1-405B-Instruct, it has been significantly customized using Neural Architecture Search (NAS), resulting in enhanced efficiency, reduced memory usage, and improved inference latency. The model supports a context length of up to 128K tokens and can operate efficiently on an 8x NVIDIA H100 node. Note: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.

ProviderNVIDIA
ReleasedApril 2025
Model IDnvidia/llama-3.1-nemotron-ultra-253b-v1

Technical specifications

Context window
131K tokens
Reasoning
Input
Text
Output
Text

Independent benchmarks

Scores below come from Artificial Analysis, an independent third party. BazaarLink does not participate in the testing.

Intelligence 9
Coding
Math 64
MMLU 83
GPQA 73
This modelCategory leader
Overall intelligence
#385out of 573 models
Aggregated across academic benchmarks
Math reasoning
#108out of 269 models
AIME / MATH-500 / SciCode
Time to first token
#461out of 586 models
AA 全球 p50:0.70 秒
Intelligence
9
Mid-tier
This model
9
Category leader
61
Median
15
Math
64
Mid-tier
This model
64
Category leader
99
Median
53
MMLU Pro
83%
This model
83%
Category leader
90%
Median
75%
GPQA
73%
This model
73%
Category leader
94%
Median
69%
LiveCodeBench
64%
This model
64%
Category leader
42%
Median
42%
HLE
8%
This model
8%
Category leader
53%
Median
7%
SciCode
35%
This model
35%
Category leader
60%
Median
33%
IFBench
38%
This model
38%
Category leader
83%
Median
44%
τ²-Bench Telecom
11%
This model
11%
Category leader
99%
Median
46%
AA-LCR
7%
This model
7%
Category leader
76%
Median
40%
Terminal-Bench Hard
2%
This model
2%
Category leader
66%
Median
14%

Top 5 — CodingCoding Index leaderboard

1GPT-5.6 Sol (xhigh)openai/gpt-5-6-sol-xhigh78$5 / $30
4GPT-5.6 Sol (high)openai/gpt-5-6-sol-high77$5 / $30
5Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)anthropic/claude-opus-5-xhigh77$5 / $25
AASource: Artificial Analysis · Independent third-party evaluation; not influenced by BazaarLinkView full evaluation on AA →
← All Models

Frequently Asked Questions

Can I still use the retired Llama 3.1 Nemotron Ultra 253b V1 model?

No. Llama 3.1 Nemotron Ultra 253b V1 has been retired and is retained here only as a historical reference. See the active alternatives linked on this page.

Relay integrity checks

No integrity-check data is available for this model yet.

BazaarLink Probe verifies endpoints that claim to serve this model and flags behavior that differs from the reference baseline.

View the full integrity report →

Related models and links

More from this provider: NVIDIA

Nemotron 3 Super 120b A12bNemotron 3 Nano 30b A3bGlm 4.7Glm 5Glm 5.2Deepseek V3.2
ModelsAPI documentationAPI latency probe
Support
Support
Hi! How can we help you?
Send a message and we'll get back to you soon.