priorsBuilder

NVIDIA · released 2026-06-04

Nemotron 3 Ultra

Measured on 31 benchmarks across 3 reasoning efforts.

General frontier

49.4

#37 of 136 models · best at none

includes estimates: 5% of the mix is measured for this model

Reasoning efforts in your mix

none49.4#82 of 313
xhigh42.2#127 of 313
max30.2#219 of 313

Your mix’s benchmarks

3 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #160max of 58728.4%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #134max of 526-0.4

Leader: Opus 5.5 · 46.4

reasoning

BullshitBench v2

196 results
  1. #77none of 19649.0%
  2. #97xhigh of 19640.0%

Leader: Opus 4.8 · 95.0%

All 31 benchmarks

show 28 more

agentic

AA-AnalystAgent

44 results
  1. #43max of 446.3%

Leader: Gemini 3.7 Flash · 60.0%

agentic

AA-Briefcase v1.1

224 results
  1. #140max of 224876

Leader: Sonnet 5.5 · 1823

agentic

AutomationBench-AA

219 results
  1. #171max of 2193.0%

Leader: Gemini 4 · 77.5%

agentic

EnterpriseOps-Gym-AA

50 results
  1. #44max of 5028.9%

Leader: Fable 5 · 51.1%

agentic

τ²-Bench Telecom

393 results
  1. #112max of 39383.3%

Leader: GLM 5.2 · 99.1%

agentic

τ³-Banking

206 results
  1. #125max of 20614.2%

Leader: Qwen 3.8 Max · 51.3%

coding

SciCode

226 results
  1. #166max of 22640.3%

Leader: Opus 5.5 · 66.9%

coding

Terminal-Bench Hard-AA

387 results
  1. #71max of 38736.4%

Leader: GPT-5.6 Sol · 65.9%

coding

Terminal-Bench v2.1-AA

238 results
  1. #111max of 23853.9%

Leader: Fable 5.1 · 91.4%

coding

Terminal-Bench v4-AA

216 results
  1. #139max of 2160.5%

Leader: Sonnet 5.5 · 63.6%

coding

WeirdML v2

127 results
  1. #77max of 12743.5%

Leader: GPT-6 Astra Pro · 93.6%

composite

AA Intelligence Index v4.3.2

610 results
  1. #188max of 61022.9

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #98max of 2201426

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #125max of 2201397

Leader: Gemini 4 · 1553

cost

AA Intelligence Index Cost per Task

177 results
  1. #96max of 177$0.60

Leader: GPT-6 Luna · $0.00

cost

Cost to Run AA Intelligence Index

177 results
  1. #86max of 177$805

Leader: GPT-6 Luna · $10.63

economics

GDP.pdf

210 results
  1. #163max of 2105.0%

Leader: Muse Wonder Full (Max) · 32.6%

economics

GDPval-AA v2.1

288 results
  1. #156max of 2881013

Leader: Opus 5.5 · 1866

instruction

IFBench

399 results
  1. #4max of 39981.4%

Leader: Grok 4.3 · 83.3%

knowledge

AA-Omniscience Accuracy

526 results
  1. #261max of 52622.6%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #40max of 52670.3%

Leader: MiniCPM5-1B (Non-reasoning) · 99.1%

knowledge

GPQA Diamond

539 results
  1. #109max of 53986.7%

Leader: GPT-6 Astra · 96.3%

long-context

AA-LCR v1.1

523 results
  1. #101max of 52379.3%

Leader: Kimi K3 · 88.7%

long-context

MLCR-AA

94 results
  1. #64max of 9411.1%

Leader: Sonnet 5.5 · 75.0%

reasoning

CritPt

529 results
  1. #174max of 5293.1%

Leader: Muse Wonder Full (Max) · 32.9%

safety

CWE-Bench

19 results
  1. #18max of 1927.5%

Leader: Grok 4.7 · 68.3%

speed

Latency (Time to First Token)

351 results
  1. #87max of 3511.1s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #84max of 351157.1

Leader: Celeris-1 · 1522.5