NEWHow AI models draw the world map · Gecko Tests
Beta
Home/Models/Nemotron 3 Ultra
NVIDIA logo

Nemotron 3 Ultra

by NVIDIA · Released Jun 2026

Open Source
44.5
avg score
Rank #179
Compare
Better than 44% of all models
Context
262K tokens (~131 books)
Input $/1M
$0.60
Output $/1M
$2.40
Type
text
License
Open Source
Benchmarks
30 tested
Data as of
About

NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...

Tested on 30 benchmarks · BenchGecko score 44.5. Top scores: Artificial Analysis · GPQA Diamond (86.7%), OTIS Mock AIME 2024-2025 (86.7%), Dtbench (83.5%).

Looking for similar performance at lower cost?
Qwen3.7 Flash scores 44.3 (100% as good) at $0.03/1M input · 95% cheaper
Capabilities
coding
43.8
#129 globally
reasoning
39.8
#109 globally
math
70.6
#47 globally
knowledge
46.6
#162 globally
agentic
22.7
#57 globally
speed
47.0
#42 globally
general
35.2
#108 globally
language
55.2
#117 globally
Benchmark Scores
Compare All
Tested on 30 benchmarks · Ranked across 8 categories
Score Distribution (all 319 models)
0255075100
▲ You are here
codingCompare coding →
LiveBench — Coding

Regularly refreshed coding problems that avoid data contamination. New problems added monthly to prevent memorization.

71.3·
LiveBench — Agentic Coding

LiveBench coding tasks that require multi-step reasoning and tool use. Tests planning and execution of complex coding workflows.

46.7·
WeirdML

Unusual and adversarial machine learning challenges. Tests robustness of reasoning about edge cases in ML systems.

43.5·
reasoningCompare reasoning →
LiveBench — Data Analysis

Fresh data analysis tasks testing ability to interpret tables, charts, and statistical data.

42.0·
LiveBench — Reasoning

Regularly refreshed reasoning problems testing logical deduction, spatial reasoning, and analytical thinking.

37.5·
mathCompare math →
OTIS Mock AIME 2024-2025

Mock AIME (American Invitational Mathematics Exam) problems from OTIS. Tests mathematical competition performance.

86.7·
LiveBench — Mathematics

Regularly updated math problems that test numerical reasoning, algebra, calculus, and combinatorics.

54.5·
Quick compare:xAI logovs Grok 3Alibaba Qwen logovs Qwen3.7 FlashAnthropic logovs Claude Sonnet 4
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Model Family · NVIDIA Nemotron 3
Dec 2025
$0.05/M in262Kctx
Dec 2025
$0.00/M in(-0.05)256Kctx(-6K)
Apr 2026
$0.00/M in256Kctx
Mar 2026
10.2
$0.08/M in(+0.08)262Kctx(+6K)18 benchmarks
Mar 2026
$0.00/M in(-0.08)262Kctx
Jun 2026
44.5+34.3
$0.60/M in(+0.60)262Kctx30 benchmarks
Jun 2026
$0.60/M in512Kctx(+250K)
Jun 2026
$0.00/M in(-0.60)1.0Mctx(+488K)
Oct 2026
n/a262Kctx(-738K)
Jun 2026
$0.20/M in131Kctx(-131K)
Jun 2026
$0.00/M in(-0.20)128Kctx(-3K)
Aug 2026
$0.07/M in(+0.07)262Kctx(+134K)7 benchmarks
Aug 2026
$0.00/M in(-0.07)1.0Mctx(+738K)
Oct 2026
n/a1.0Mctx
See the full Nemotron 3 family →
Similar Models
xAI logo
Grok 3
xAI
44.5n/a
Alibaba Qwen logo
Qwen3.7 Flash
Alibaba Qwen
44.3$0.03/1M
Anthropic logo
Claude Sonnet 4
Anthropic
44.3$3.00/1M
Links
Info
NVIDIAPricing explorerDevelopers · API
Research
Technical Report
Documentation
API DocsPlayground
Community
@NVIDIA
Source Code
GitHubHugging Face
BenchGecko API
nemotron-3-ultra-550b-a55b
Specifications
  • Typetext
  • Context262K tokens (~131 books)
  • ReleasedJun 2026
  • LicenseOpen Source
  • StatusActive
  • Cost / Message~$0.004
Available On
NVIDIA logoNVIDIA$0.60
Categories
codingreasoningmathknowledgeagenticspeedgenerallanguage
Learn More
context-windowtransformeropen weightstokens
Share & Export
Tweet
Related Models
xAI logoGrok 3
44.5
Alibaba Qwen logoQwen3.7 Flash
44.3
Anthropic logoClaude Sonnet 4
44.3
Google DeepMind logoGemini 2.0 Flash
44.3
Nemotron 3 Ultra is an open-source text AI model by NVIDIA, released in June 2026. It has an average benchmark score of 44.5. Context window: 262K tokens.

Related Models

Grok 3 · xAIQwen3.7 Flash · Alibaba QwenClaude Sonnet 4 · AnthropicGemini 2.0 Flash · Google DeepMindGemini 2.0 Flash (Dec 2024) · Google DeepMind

Benchmarks

Artificial Analysis · GPQA DiamondOTIS Mock AIME 2024-2025DtbenchArtificial Analysis · tau2-Bench TelecomArtificial Analysis · IFBench

Related Pages

NVIDIA · ProviderNVIDIA · EconomyAll ModelsCompare ModelsPricingDevelopers · APIContext Window · Glossary

Key facts · as of 2026-10-16

  • Nemotron 3 Ultra by NVIDIA. BenchGecko score 44.5, rank 178 of 319 scored models (normalized average of public benchmark scores).
  • List price $0.60 input · $2.40 output per 1M tokens (as of 2026-10-16).
  • Sold by 5 providers (as of 2026-10-16): DeepInfra (fp4) $0.40 in / $1.76 out · DeepInfra (fp4) $0.50 in / $2.20 out · BaseTen (fp4) $0.60 in / $2.40 out · BaseTen (fp4) $0.60 in / $2.40 out · Venice (fp8) $0.63 in / $3.13 out. Every provider

How to cite · data as of 2026-10-16

Nemotron 3 Ultra · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-16. https://benchgecko.ai/model/nemotron-3-ultra-550b-a55b

Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP