Nemotron 3 Ultra 550B A55B
Nemotron 3 Ultra 550B A55B is at the frontier in instruction following; strong in long context; and behind the leaders in factuality, reasoning, coding, math, and agentic tasks. Too few results yet to rate safety, multimodal tasks, or multilingual tasks.
Capability profile
Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.
Too few results yet to rate Safety, Multimodal or Multilingual.
Price
$0.60input$2.50outputper million tokens
From Artificial Analysis · 4 providers tracked · All prices
Evidence
109results on70benchmarks
- 12 independently verified
- 18 aggregator
- 60 vendor-reported
- 19 cross-referenced
From 12 sources · latest Oct 8, 2026 · How verification works
API features
Tool callingReasoning
As listed by OpenRouter
Nemotron 3 Ultra 550B A55B benchmark results
109 results on 70 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.
5.5% behind the leader3 of 3 ranked benchmarks measured
- IFBench81.36Oct 8, 2026aa_ifbench
- LiveBench · Instruction Following73.45Oct 8, 2026livebench_instruction_following@2026-06-25
- 63.80Oct 7, 2026
6.5% behind the leader2 of 3 ranked benchmarks measured
- 61.90Oct 7, 2026
- 79.33Oct 8, 2026
Show 1 more long context resultHide 1 long context result
- 65.40Oct 7, 2026
29.6% behind the leader3 of 4 ranked benchmarks measured
- AA-Omniscience · Non-hallucination70.31Oct 8, 2026omniscienceNonHallucination
- 32.40Aug 24, 2026
- AA-Omniscience · Accuracy22.58Oct 8, 2026omniscienceAccuracy
31.9% behind the leader5 of 6 ranked benchmarks measured
- GPQA Diamond86.67Oct 8, 2026gpqa
- LiveBench · Reasoning74.70Oct 8, 2026livebench_reasoning@2026-06-25
- 41.70Jun 8, 2026
- Humanity's Last Exam28.41Oct 8, 2026aa_hle
- 3.14Oct 8, 2026
Show 9 more reasoning resultsHide 9 reasoning results
- 3.10Oct 7, 2026
- 3.10Jul 30, 2026
- GPQA Diamond87.00Oct 7, 2026GPQA
- GPQA Diamond87.90Sep 1, 2026GPQA (no tools)
- 86.70Jul 30, 2026
- 37.40Oct 7, 2026
- Humanity's Last Exam26.10Sep 1, 2026HLE (no tools)
- Humanity's Last Exam26.70Jun 6, 2026HLE (no tools)
- Humanity's Last Exam26.60Jul 30, 2026HLE text only
38.4% behind the leader8 of 10 ranked benchmarks measured
- 89.00Oct 7, 2026
- LiveBench · Coding70.70Oct 8, 2026livebench_coding@2026-06-25
- 70.70Oct 7, 2026
- 67.70Oct 7, 2026
- SciCode40.28Oct 8, 2026aa_scicode
- Terminal-Bench 2.153.93Oct 8, 2026terminalbenchV21
- Terminal-Bench Hard36.36Oct 8, 2026aa_terminalbench_hard
- LiveBench · Agentic Coding38.74Oct 8, 2026livebench_agentic_coding@2026-06-25
Show 8 more coding resultsHide 8 coding results
- 44.60Oct 7, 2026
- 39.90Jul 30, 2026
- 69.10Sep 1, 2026
- 69.50Sep 1, 2026
- 70.70Jul 30, 2026
- 56.40Oct 7, 2026
- 53.90Sep 1, 2026
- 56.40Jul 30, 2026
38.7% behind the leader1 of 5 ranked benchmarks measured
- LiveBench · Mathematics88.66Oct 8, 2026livebench_math@2026-06-25
Show 3 more math resultsHide 3 math results
- 94.20Jul 30, 2026
- 78.80Jul 30, 2026
- 92.30Oct 7, 2026
43.8% behind the leader6 of 7 ranked benchmarks measured
- 63.10Oct 8, 2026
- 44.40Oct 7, 2026
- 25.81Oct 8, 2026
- τ-Bench V3 · Banking14.23Oct 8, 2026tauBanking
- Terminal-Bench 4.00.51Oct 8, 2026
Show 7 more agentic resultsHide 7 agentic results
- 41.40Sep 1, 2026
- GDPval (win rate)47.90Sep 1, 2026GDPVal
- GDPval (win rate)46.70Jun 6, 2026GDPVal
- 62.50Sep 18, 2026
- 44.70Jul 16, 2026
- 19.20Sep 1, 2026
- 22.60Jun 6, 2026
More results
Benchmarks outside the capability baskets. They are not ranked against a leader.
- livebench_data_analysis54.46Oct 8, 2026livebench_data_analysis@2026-06-25
- livebench_language70.81Oct 8, 2026livebench_language@2026-06-25
- AA Intelligence23.00Sep 21, 2026Artificial Analysis Intelligence Index
- AA Intelligence38.00Jun 24, 2026Artificial Analysis Intelligence Index
- τ²-Bench Telecom (AA run)83.33Oct 8, 2026aa_tau2
- AA Intelligence22.93Oct 8, 2026aa_intelligence_index
Show 46 more resultsHide 46 results
- 21.68Sep 9, 2026
- -0.40Oct 8, 2026
- Apex (Pass@1)84.80Oct 7, 2026Apex
- 74.90Jun 6, 2026
- 84.80Jun 6, 2026
- Artificial Analysis Coding Index49.27Sep 9, 2026aa_coding_index
- BrowseComp (context management)63.00Aug 24, 2026BrowseComp (with context management)
- 3.40Sep 1, 2026
- 3.10Jun 6, 2026
- 53.70Oct 7, 2026
- 37.53Oct 7, 2026
- 1164.00Jul 30, 2026
- 85.60Aug 24, 2026
- 37.40Jun 6, 2026
- HLE (with tools)37.40Jul 30, 2026HLE with tools
- 81.70Jun 10, 2026
- 82.30Sep 1, 2026
- 92.30Jun 6, 2026
- 564.70Sep 1, 2026
- 570.00Jun 6, 2026
- 86.80Oct 7, 2026
- 83.00Oct 7, 2026
- 83.00Jun 10, 2026
- 90.00Oct 7, 2026
- 89.80Sep 1, 2026
- 56.40Sep 1, 2026
- 56.00Jun 6, 2026
- 94.70Oct 7, 2026
- 94.00Sep 1, 2026
- 43.50Sep 1, 2026
- 44.60Jun 6, 2026
- 98.70Aug 24, 2026
- 46.40Jul 30, 2026
- 80.00Sep 1, 2026
- 81.50Jun 6, 2026
- 70.30Sep 1, 2026
- 70.90Jun 6, 2026
- 88.40Sep 1, 2026
- 86.40Jun 6, 2026
- 93.60Sep 1, 2026
- TauBench V3 - Telecom92.90Aug 24, 2026Telecom
- 34.30Jul 30, 2026
- 53.70Jun 6, 2026
- 60.10Jun 6, 2026
- 83.70Jun 10, 2026
- τ³-Bench Banking13.80Jul 30, 2026τ³-Banking
Nemotron 3 Ultra 550B A55B: common questions
Who makes Nemotron 3 Ultra 550B A55B?
Nemotron 3 Ultra 550B A55B is made by NVIDIA.
When was Nemotron 3 Ultra 550B A55B released?
Nemotron 3 Ultra 550B A55B was released on Jun 4, 2026, according to Artificial Analysis.
What is Nemotron 3 Ultra 550B A55B good at?
Nemotron 3 Ultra 550B A55B is at the frontier in instruction following; strong in long context; and behind the leaders in factuality, reasoning, coding, math, and agentic tasks. Too few results yet to rate safety, multimodal tasks, or multilingual tasks.
How much does Nemotron 3 Ultra 550B A55B cost?
Nemotron 3 Ultra 550B A55B costs $0.60 per million input tokens and $2.50 per million output tokens, according to Artificial Analysis. We track its price at 4 providers. At a mix of three input tokens to one output token, it costs more than 60% of the 330 priced models we track.
How many benchmarks has Nemotron 3 Ultra 550B A55B been tested on?
We track 109 results for Nemotron 3 Ultra 550B A55B on 70 benchmarks from 12 sources, 12 of them independently verified. The latest was recorded on Oct 8, 2026.
Which API features does Nemotron 3 Ultra 550B A55B support?
OpenRouter lists tool calling and reasoning for Nemotron 3 Ultra 550B A55B.
About this record
Where Nemotron 3 Ultra 550B A55B's numbers come from, and every name it appears under.
- Tracked since
- Jun 24, 2026
- Newest source mention
- Oct 8, 2026
Where the results come from
Verification: 109 scores · 12 independently verified · 18 aggregator-attributed · 19 vendor cross-reference · 60 vendor-reported. How these tiers are assigned
From 12 sources on 7 sites. Hugging Face supplies 45 of them; the 12 independently verified results come from 4 sites. Bars are coloured by trust tier.
- huggingface.co45
- api.llm-stats.com21
- artificialanalysis.ai20
- thinkingmachines.ai13
- livebench.ai7
- labs.scale.com2
- simple-bench.com1
Also known as
How our sources name Nemotron 3 Ultra 550B A55B at each reasoning setting.
| Setting | Short form |
|---|---|
| high | nemotron-3-ultra-nvfp4 high |
| xhigh | nemotron 3 ultra (xhigh) |