RankingEffort benchmarks

Amazon Nova 2 Pro Preview effort benchmarks

Compare measured Low, Medium, High and other settings. Every number keeps its source.

Model

Amazon Nova 2 Pro Preview

47 measured results · 3 reported effort labels · 17 benchmark/harness combinations. Model evidence and sources →

Capability score evidence

Amazon Nova 2 Pro Preview · Medium: 9.3 overall comparison score. Meets the overall evidence requirements.

Which sources carry the weight?

Artificial Analysis · 100.0% direct comparison weight · independent

These shares describe the comparisons entering the fit, not fractions of the final score.

How sensitive is this score?

Remove one evaluator or family and refit the graph against the same overall panel. Values use the matching dataset audited 2026-09-12. These are sensitivity checks, not confidence intervals; ranking eligibility may also change.

Removed evidence	Recomputed score	Change
Artificial Analysis	Comparison path lost	—
ARC Prize	11.1	+1.8
Datacurve	8.0	-1.3
LiveCodeBench	9.2	-0.1
Scale	10.2	+0.9
Terminal-Bench	9.3	-0.0
Cognition	9.7	+0.4
LiveBench	9.5	+0.2
SWE-rebench	9.5	+0.2
Epoch AI	9.3	-0.0
Vals AI	9.0	-0.3
Provider reports · Anthropic	9.3	+0.0
Provider reports · Moonshot AI	9.3	-0.0
Provider reports · Meta	9.3	-0.0
Provider reports · Mistral	9.3	+0.0
All provider reports	9.3	-0.0

Every benchmark family

Removed evidence	Recomputed score	Change
arc-agi	11.0	+1.7
deepswe	8.0	-1.3
ifbench	8.2	-1.1
mmmu-pro	10.4	+1.1
critpt	10.3	+1.0
hle	10.2	+0.9
tau-bench	8.7	-0.6
epoch-game-puzzles	8.8	-0.4
frontiercode	9.7	+0.4
simpleqa	9.6	+0.3
aime	9.0	-0.3
omniscience	9.0	-0.3
gdpval	9.5	+0.3
swe-rebench	9.5	+0.2
gpqa	9.1	-0.2
aa-lcr	9.1	-0.2
scicode	9.4	+0.1
multi-swe-bench	9.2	-0.1
livebench-instructions	9.4	+0.1
mmlu-pro	9.4	+0.1
terminal-bench	9.4	+0.1
harvey	9.2	-0.1
livebench-language	9.4	+0.1
apex-agents	9.4	+0.1
gdp-pdf	9.4	+0.1
swe-atlas-qna	9.2	-0.1
frontiermath	9.4	+0.1
livebench-coding	9.2	-0.1
gmmlu	9.4	+0.1
arc-agi-3	9.2	-0.1
vibe-code	9.2	-0.1
livebench-reasoning	9.2	-0.1
livecodebench	9.2	-0.1
vals-finance-agent	9.3	-0.0
math500	9.3	+0.0
briefcase	9.3	-0.0
vals-legal-research	9.3	-0.0
swe-atlas-test-writing	9.3	-0.0
vals-code-migration	9.3	-0.0
automationbench	9.3	-0.0
analyst-agent	9.3	+0.0
livebench-data	9.3	+0.0
itbench	9.3	-0.0
livebench-math	9.3	+0.0
vals-excel-modeling	9.3	-0.0
swe-atlas-refactoring	9.3	-0.0
enigma-eval	9.3	+0.0
enterprise-ops	9.3	-0.0
swe-bench-pro	9.3	+0.0

What evidence is missing from the fit?

Across all collected settings for this model: 66 observations, 44 contributing, 0 matched without graph weight and 22 excluded. Counts do not establish rank eligibility.

3 · Benchmark family or source protocol has not been reviewed
11 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately
8 · Supporting evidence outside the reviewed capability core

Inspect every observation and exclusion →

Capability profile

One configuration, seven capabilities

Comparison score · 0–100

AAmazon Nova 2 Pro Preview · Medium4 of 7 capabilities supported

Gaps are unknown, not zero. Hollow points are preliminary.

Agentic
A18.9Supported
Hard reasoning
A7.8Supported
Coding
A9.0Supported
Human pref
Unknown
Knowledge
A20.1Supported
Multimodal
A6.8Preliminary
Long context
A15.8Preliminary

Each profile belongs to the named configuration. Capability scores use a shared panel for all axes, distinct from the overall ranking panel; they are not accuracy percentages or direct head-to-head wins.

Compare this configuration →Explore the model evidence profile →

These are published benchmark scores. Supported effort estimates also appear on the Capability leaderboard. Results from different harnesses occupy different rows. Missing cells stay unknown; no scores are copied between effort levels.

What these comparisons mean

Each source reports its own effort labels. Matching labels do not establish equal compute budgets or identical fallback behavior. Open Reported settings to inspect the exact model name, agent and task coverage. Multiple results at one level remain visible rather than selecting the best. Source-reported None or Non-reasoning labels are not treated as evidence that an API supports those settings.

These measurements use their own dated source collection. The leaderboard fits identified configurations jointly; it contains no pooled model ratings. Explicit thinking budgets are separate configurations. Read the methodology.

Download all effort benchmark results · 7,424 results across 240 models · Collected 2026-09-12

Benchmark / source	Low	Medium	Non-reasoning
AIME 2025Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	63.33% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	89% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	30.67% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
AA-LCR v1.1Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	64.67% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	64% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	30% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
CritPtArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by Argonne and UIUC, with contributions from 60+ researchers globally · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	0% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	0% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	0% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
GDPval-AA v2 LeaderboardArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Elo rating for performance on real-world work tasks · Anchored to a human baseline of 1,000 · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-09-12.	600.77 Elo Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	629.57 Elo Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	508.59 Elo Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
Global-MMLU-LiteArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	87.6% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	Not reported	78.63% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
GPQA DiamondArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	75.05% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	78.48% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	63.64% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
Humanity's Last ExamArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	5.19% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	9.41% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	3.94% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
IFBenchArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	79.59% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	79.05% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	52.04% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
LiveCodeBenchArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	63.81% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	73.02% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	47.3% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
MLCR-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of tasks judged accurate, complete, and concise · Expert + compound (hard) sets · Benchmark developed by Wisedocs · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	Not reported	3.89% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	Not reported
MMLU-ProArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	82.16% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	82.95% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	77.16% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
MMMU-ProArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	62.72% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	64.51% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	Not reported
AA-Omniscience IndexArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct. · Evaluation results measured independently by Artificial Analysis Unit: index. Source collected 2026-09-12.	-40.53 Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	-49.05 Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	-49.5 Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
AA-Omniscience AccuracyArtificial Analysis · 6000 questions; 1 attempt; correct-answer share; GPT-5.6 Luna medium grader Benchmark details AA-Omniscience Accuracy (higher is better) measures the proportion of correctly answered questions out of all questions, regardless of whether the model chooses to answer · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	17.1% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	21.97% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	16.92% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
𝜏²-Bench TelecomArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	90.64% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	92.69% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	71.64% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
Terminal-Bench HardArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	17.42% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	24.24% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	16.67% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro
Terminal-Bench v2.1Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by the Laude Institute and open-source contributors · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	19.48% Reported settings Nova 2.0 Pro Preview (low); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-low	29.59% Reported settings Nova 2.0 Pro Preview (medium); Model source: https://artificialanalysis.ai/models/nova-2-0-pro-reasoning-medium	17.23% Reported settings Nova 2.0 Pro Preview (Non-reasoning); Model source: https://artificialanalysis.ai/models/nova-2-0-pro

Amazon Nova 2 Pro Preview effort benchmarks