Benchmark Almanac

Nova 2.0 Pro Preview

Nova 2.0 Pro Preview ranks first on 2 of the 16 benchmarks with five or more models scored.

Amazon Released December 2nd, 2025 Compare with others
21
Benchmarks with a score
2
Ranked first, out of five or more models
5
In the top three
Amazon Dec 2025

Nova 2.0 Pro Preview

Closed weights

145.3 #129 of 404
Self-Reported Index, from the lab's own numbers. No Epoch Capabilities Index score yet.
  • MMMU-Pro 63.5% #62 /90
  • IFBench 80.2% #7 /80
  • BFCL V4 61.6% #18 /30

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

BenchmarkScoreRankSource
MultiChallenge 77.7% #1 of 42 Lab model card
IFBench 80.2% #7 of 80 Lab model card
MMLU-Pro 81.6% #65 of 167 Lab model card
LongCodeBench 84% #1 of 2 Lab model card

Math and science

BenchmarkScoreRankSource
AIME 2025 92.3% #40 of 143 Lab model card
GPQA Diamond 81.4% #109 of 302 Lab model card

Coding

BenchmarkScoreRankSource
LiveCodeBench v5 74.6% #3 of 47 Lab model card
SWE-bench Verified 70% #78 of 155 Lab model card

Agents and tools

BenchmarkScoreRankSource
tau2-bench 92.7% #10 of 79 Lab model card
Terminal-Bench 41.3% #9 of 22 Lab model card
ScreenSpot 88.1% #5 of 11 Lab model card
BFCL V4 61.6% #18 of 30 Lab model card
tau2-bench Verified 71.5% #2 of 3 Lab model card

Vision, audio and robotics

BenchmarkScoreRankSource
Common Voice 5.50% #1 of 5 Lab model card
OCRBench v2 64.5% #7 of 22 Lab model card
RealKIE-FCC Verified 67% #1 of 3 Lab model card
QVHighlights 76.7% #2 of 5 Lab model card
MMMU-Pro 63.5% #62 of 90 Lab model card

Safety and security

BenchmarkScoreRankSource
WildGuardTest 84.9% #3 of 5 Lab model card
Aegis 2.0 81.4% #2 of 3 Lab model card
Jigsaw Toxic Comment Classification 54.7% #3 of 3 Lab model card

Its lab's numbers put it at 145.3, number 129 of 404

The Self-Reported Index fits one score to the 13 capability results Amazon reported for this model, on the scale of the Epoch Capabilities Index. Labs choose which benchmarks to show and how to run them, so read it as the lab's claim, not an independent measurement. The 90% interval runs from 142.2 to 148.5. How we compute it.

Its launch post

The benchmarks the lab chose to headline, with the scores it gave.

Amazon Nova 2: Multimodal Reasoning and Generation Models (Technical Report and Model Card) Dec 2, 2025

Its system card

Every result in the longer documents the lab published with it. 4 of them aren't in the launch post.