BenchAtlas

Rankings / OpenAI

GPT 5.4 Nano (2026-03-17)

proprietary

released 2026-03-17

BenchAtlas Index

as of 2026-07-14
52.6
high effort
rank #199
6 families · 3 categories · medium
51.6
xhigh effort
rank #206
13 families · 5 categories · high
47.9
medium effort
rank #235
8 families · 5 categories · high

Benchmark evidence

110 results
agentic coding
Release 2026-06-25
tasks_counted=3 · livebench_version=2026-06-25
xhigh effort46.8 %independentLiveBench
τ²-Benchxhigh effort76.0 %independentArtificial Analysis
τ²-Benchmedium effort52.6 %independentArtificial Analysis
τ²-Benchno reasoning34.8 %independentArtificial Analysis
τ²-Bench
subset=banking
xhigh effort21.0 %independentArtificial Analysis
coding
Release 2026-06-25
tasks_counted=2 · livebench_version=2026-06-25
xhigh effort70.8 %independentLiveBench
SciCodexhigh effort46.9 %independentArtificial Analysis
SciCodemedium effort38.4 %independentArtificial Analysis
SciCodeno reasoning35.2 %independentArtificial Analysis
data analysis
Release 2026-06-25
tasks_counted=3 · livebench_version=2026-06-25
xhigh effort67.6 %independentLiveBench
external indices
AA Coding Indexxhigh effort56.1 pointsindependentArtificial Analysis
Intelligence Index v4.1xhigh effort38.2 pointsindependentArtificial Analysis
Intelligence Index v4.1medium effort30.2 pointsindependentArtificial Analysis
Intelligence Index v4.1no reasoning17.6 pointsindependentArtificial Analysis
ECIxhigh effort146.6 pointsindependentEpoch AI Benchmarking Hub
ECI146.6 pointsindependentEpoch AI Benchmarking Hub
ECImedium effort146.6 pointsindependentEpoch AI Benchmarking Hub
ECIhigh effort146.6 pointsindependentEpoch AI Benchmarking Hub
ECIlow effort146.6 pointsindependentEpoch AI Benchmarking Hub
ECIno reasoning146.6 pointsindependentEpoch AI Benchmarking Hub
Vals Indexhigh effort46.6 pointsindependentVals AI
factuality
SimpleQA Verifiedhigh effort12.0 %independentEpoch AI Benchmarking Hub
2026-04-14
human preference
Coding (style control)older version
arena=text · category=coding · style_control=true
high effort1461.2
1454.9–1467.6
communityLMArena Leaderboard Dataset
Industry Software And It Services (style control)older version
arena=text · category=industry_software_and_it_services · style_control=true
high effort1451.2
1445.6–1456.8
communityLMArena Leaderboard Dataset
Expert (style control)older version
arena=text · category=expert · style_control=true
high effort1436.7
1427.3–1446.2
communityLMArena Leaderboard Dataset
Industry Mathematical (style control)older version
arena=text · category=industry_mathematical · style_control=true
high effort1436.6
1424.4–1448.8
communityLMArena Leaderboard Dataset
Hard Prompts English (style control)older version
arena=text · category=hard_prompts_english · style_control=true
high effort1432.3
1426.3–1438.3
communityLMArena Leaderboard Dataset
French (style control)older version
arena=text · category=french · style_control=true
high effort1428.6
1411.7–1445.6
communityLMArena Leaderboard Dataset
Math (style control)older version
arena=text · category=math · style_control=true
high effort1426.9
1415.2–1438.6
communityLMArena Leaderboard Dataset
Hard Prompts (style control)older version
arena=text · category=hard_prompts · style_control=true
high effort1423.5
1418.6–1428.4
communityLMArena Leaderboard Dataset
Chinese (style control)older version
arena=text · category=chinese · style_control=true
high effort1422.0
1409.3–1434.7
communityLMArena Leaderboard Dataset
Industry Medicine And Healthcare (style control)older version
arena=text · category=industry_medicine_and_healthcare · style_control=true
high effort1419.1
1408.5–1429.8
communityLMArena Leaderboard Dataset
Industry Life And Physical And Social Science (style control)older version
arena=text · category=industry_life_and_physical_and_social_science · style_control=true
high effort1419.0
1411.5–1426.6
communityLMArena Leaderboard Dataset
English (style control)older version
arena=text · category=english · style_control=true
high effort1415.9
1410.7–1421.2
communityLMArena Leaderboard Dataset
Multi Turn (style control)older version
arena=text · category=multi_turn · style_control=true
high effort1415.6
1408.4–1422.9
communityLMArena Leaderboard Dataset
Longer Query (style control)older version
arena=text · category=longer_query · style_control=true
high effort1405.7
1399.9–1411.5
communityLMArena Leaderboard Dataset
Industry Business And Management And Financial Operations (style control)older version
arena=text · category=industry_business_and_management_and_financial_operations · style_control=true
high effort1405.2
1398.1–1412.3
communityLMArena Leaderboard Dataset
Industry Legal And Government (style control)older version
arena=text · category=industry_legal_and_government · style_control=true
high effort1404.0
1393.7–1414.3
communityLMArena Leaderboard Dataset
Overall (style control)
arena=text · category=overall · style_control=true
high effort1403.7
1399.7–1407.7
communityLMArena Leaderboard Dataset
Polish (style control)older version
arena=text · category=polish · style_control=true
high effort1399.6
1379.4–1419.8
communityLMArena Leaderboard Dataset
Russian (style control)older version
arena=text · category=russian · style_control=true
high effort1393.4
1384.5–1402.3
communityLMArena Leaderboard Dataset
German (style control)older version
arena=text · category=german · style_control=true
high effort1392.3
1369.2–1415.4
communityLMArena Leaderboard Dataset
Exclude Ties (style control)older version
arena=text · category=exclude_ties · style_control=true
high effort1391.4
1386.1–1396.8
communityLMArena Leaderboard Dataset
Instruction Following (style control)older version
arena=text · category=instruction_following · style_control=true
high effort1388.1
1382.2–1394.0
communityLMArena Leaderboard Dataset
Non English (style control)older version
arena=text · category=non_english · style_control=true
high effort1387.8
1382.7–1392.8
communityLMArena Leaderboard Dataset
Spanish (style control)older version
arena=text · category=spanish · style_control=true
high effort1385.8
1368.9–1402.7
communityLMArena Leaderboard Dataset
Industry Writing And Literature And Language (style control)older version
arena=text · category=industry_writing_and_literature_and_language · style_control=true
high effort1364.6
1358.0–1371.2
communityLMArena Leaderboard Dataset
Industry Entertainment And Sports And Media (style control)older version
arena=text · category=industry_entertainment_and_sports_and_media · style_control=true
high effort1352.1
1345.0–1359.3
communityLMArena Leaderboard Dataset
Korean (style control)older version
arena=text · category=korean · style_control=true
high effort1338.9
1316.4–1361.5
communityLMArena Leaderboard Dataset
Creative Writing (style control)older version
arena=text · category=creative_writing · style_control=true
high effort1337.2
1329.4–1345.0
communityLMArena Leaderboard Dataset
knowledge science
GPQA Diamond
implementation=artificial-analysis
xhigh effort81.7 %independentArtificial Analysis
GPQA Diamondhigh effort78.5 %independentEpoch AI Benchmarking Hub
2026-04-14
GPQA Diamond
implementation=vals-ai
high effort77.5 %independentVals AI
GPQA Diamond
implementation=artificial-analysis
medium effort76.1 %independentArtificial Analysis
GPQA Diamond
implementation=artificial-analysis
no reasoning55.8 %independentArtificial Analysis
Humanity's Last Exam
implementation=artificial-analysis
xhigh effort26.5 %independentArtificial Analysis
Humanity's Last Exam
implementation=artificial-analysis
medium effort14.7 %independentArtificial Analysis
Humanity's Last Exam
implementation=artificial-analysis
no reasoning4.2 %independentArtificial Analysis
MMLU-Pro
implementation=vals-ai
high effort77.2 %independentVals AI
language
Release 2026-06-25
tasks_counted=3 · livebench_version=2026-06-25
xhigh effort62.5 %independentLiveBench
long context instruction
AA-LCRxhigh effort66.0 %independentArtificial Analysis
AA-LCRmedium effort57.3 %independentArtificial Analysis
AA-LCRno reasoning24.7 %independentArtificial Analysis
IFBenchxhigh effort75.9 %independentArtificial Analysis
IFBenchmedium effort64.4 %independentArtificial Analysis
IFBenchno reasoning32.7 %independentArtificial Analysis
Release 2026-06-25
tasks_counted=4 · livebench_version=2026-06-25
xhigh effort67.2 %independentLiveBench
multimodal
MMMU
implementation=vals-ai
high effort73.6 %independentVals AI
professional
CaseLawhigh effort51.9 %independentVals AI
CorpFinhigh effort61.2 %independentVals AI
LegalBenchhigh effort77.9 %independentVals AI
TaxEvalhigh effort67.4 %independentVals AI
reasoning math
AIME
implementation=vals-ai
high effort88.8 %independentVals AI
ARC-AGI-1older version
split=public_eval · model_type=CoT
high effort51.6 %independentARC Prize Leaderboard
ARC-AGI-1older version
split=semi_private · model_type=CoT
xhigh effort51.5 %independentARC Prize Leaderboard
ARC-AGI-1older version
split=public_eval · model_type=CoT
xhigh effort47.9 %independentARC Prize Leaderboard
ARC-AGI-1older version
split=public_eval · model_type=CoT
medium effort43.4 %independentARC Prize Leaderboard
ARC-AGI-1older version
split=semi_private · model_type=CoT
high effort38.2 %independentARC Prize Leaderboard
ARC-AGI-1older version
split=semi_private · model_type=CoT
medium effort33.0 %independentARC Prize Leaderboard
ARC-AGI-1older version
split=public_eval · model_type=CoT
low effort24.6 %independentARC Prize Leaderboard
ARC-AGI-1older version
split=semi_private · model_type=CoT
low effort18.3 %independentARC Prize Leaderboard
ARC-AGI-2
split=semi_private · model_type=CoT
xhigh effort5.7 %independentARC Prize Leaderboard
ARC-AGI-2
split=public_eval · model_type=CoT
high effort5.1 %independentARC Prize Leaderboard
ARC-AGI-2
split=public_eval · model_type=CoT
xhigh effort4.2 %independentARC Prize Leaderboard
ARC-AGI-2
split=semi_private · model_type=CoT
high effort3.6 %independentARC Prize Leaderboard
ARC-AGI-2
split=public_eval · model_type=CoT
medium effort3.5 %independentARC Prize Leaderboard
ARC-AGI-2
split=semi_private · model_type=CoT
medium effort1.9 %independentARC Prize Leaderboard
ARC-AGI-2
split=semi_private · model_type=CoT
low effort1.5 %independentARC Prize Leaderboard
ARC-AGI-2
split=public_eval · model_type=CoT
xhigh effort0.2 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-2
split=semi_private · model_type=CoT
xhigh effort0.2 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-2
split=public_eval · model_type=CoT
high effort0.1 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-2
split=semi_private · model_type=CoT
high effort0.1 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-1older version
split=semi_private · model_type=CoT
xhigh effort0.1 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-1older version
split=public_eval · model_type=CoT
xhigh effort0.1 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-1older version
split=semi_private · model_type=CoT
high effort0.1 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-2
split=public_eval · model_type=CoT
medium effort0.1 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-2
split=semi_private · model_type=CoT
medium effort0.1 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-1older version
split=public_eval · model_type=CoT
high effort0.1 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-1older version
split=public_eval · model_type=CoT
medium effort0.0 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-1older version
split=semi_private · model_type=CoT
medium effort0.0 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-2
split=public_eval · model_type=CoT
low effort0.0 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-1older version
split=public_eval · model_type=CoT
low effort0.0 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-1older version
split=semi_private · model_type=CoT
low effort0.0 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-2
split=semi_private · model_type=CoT
low effort0.0 usd_per_taskindependentARC Prize Leaderboard
ARC-AGI-2
split=public_eval · model_type=CoT
low effort0.0 %independentARC Prize Leaderboard
FrontierMathhigh effort25.9 %independentEpoch AI Benchmarking Hub
2026-04-15
FrontierMath Tier 4older versionhigh effort6.3 %independentEpoch AI Benchmarking Hub
2026-04-15
Release 2026-06-25
tasks_counted=4 · livebench_version=2026-06-25
xhigh effort91.0 %independentLiveBench
Release 2026-06-25
tasks_counted=4 · livebench_version=2026-06-25
xhigh effort81.1 %independentLiveBench
OTIS Mock AIME 2024–2025high effort87.8 %independentEpoch AI Benchmarking Hub
2026-04-14

Agent + model results

systems, not bare-model scores
agent + model Artificial Analysis harness + GPT 5.4 Nano (2026-03-17)Terminal-Bench 2.160.7 %independentArtificial Analysis
agent + model Artificial Analysis harness + GPT 5.4 Nano (2026-03-17)Terminal-Bench Hard42.4 %independentArtificial Analysis
agent + model Artificial Analysis harness + GPT 5.4 Nano (2026-03-17)Terminal-Bench Hard33.3 %independentArtificial Analysis
agent + model Artificial Analysis harness + GPT 5.4 Nano (2026-03-17)Terminal-Bench Hard24.2 %independentArtificial Analysis

These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.