RankingEffort benchmarks

GPT-5.4 Mini · XHigh

Compare models

92 measured results · 6 effort settings · 60 benchmark/harness combinations

Model evidence and sources →

Change model

Model

Current configurationSources

Compare with

Choose a second model and its effort to see both profiles and benchmark differences below.

Compare withModel and configuration

Capability profile

One configuration, seven capabilities

Comparison score · 0–100

AGPT-5.4 Mini · XHigh4 of 7 capabilities supported

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

Agentic
A34.1Supported
Hard reasoning
A22.5Supported
Coding
A23.8Supported
Human pref
Unknown
Knowledge
A15.6Supported
Multimodal
A18.1Preliminary
Long context
A23.3Preliminary

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.

Capability score evidence

GPT-5.4 Mini · XHigh: 13.0 overall comparison score. Meets the overall evidence requirements.

6 areas measured. The available task mix affects this estimate; missing areas remain unknown.

Which sources carry the weight?

Artificial Analysis · 21.7% direct comparison weight · independent of the model provider
Cognition · 21.3% direct comparison weight · independent of the model provider
ARC Prize · 19.7% direct comparison weight · independent of the model provider
LiveBench · 18.7% direct comparison weight · independent of the model provider
Vals AI · 18.5% direct comparison weight · independent of the model provider

Which benchmark families carry the weight?

FrontierCode · 21.3% direct comparison weight
ARC-AGI · 19.7% direct comparison weight
livebench-data · 2.7% direct comparison weight
livebench-language · 2.7% direct comparison weight
multi-swe-bench · 2.7% direct comparison weight
Other families (30) · 50.9% direct comparison weight

These shares describe direct comparison weight, not fractions of the final score or confidence. Source balancing does not give every family equal weight or remove correlations between overlapping sources.

How sensitive is this score?

The audit removes one source at a time and refits against the same overall reference panel. These published results use the matching dataset audited 2026-10-20.

Both ranks use the same surviving configurations for each removal. The cohort starts with 444 configurations supported in the original audit, including separate effort settings. These are diagnostic positions, not the default leaderboard ranks. Eligibility is not rechecked after removal.

What changes when a source is removed?
Removed source	Rank before → after	Rank movement	Cohort remaining / lost support	Score after (change)
Artificial Analysis	139 → 141	Down 2	203 remaining / 241 lost	11.4 (-1.7 points)
ARC Prize	173 → 179	Down 6	444 remaining / 0 lost	15.2 (+2.2 points)
Datacurve	173 → 172	Up 1	444 remaining / 0 lost	10.4 (-2.6 points)
LiveCodeBench	173 → 173	No movement	444 remaining / 0 lost	13.0 (+0.0 points)
Scale	173 → 173	No movement	444 remaining / 0 lost	14.6 (+1.5 points)
Terminal-Bench	173 → 173	No movement	444 remaining / 0 lost	12.9 (-0.1 points)
Cognition	173 → 162	Up 11	444 remaining / 0 lost	14.5 (+1.5 points)
LiveBench	172 → 174	Down 2	443 remaining / 1 lost	12.6 (-0.4 points)
SWE-rebench	173 → 174	Down 1	444 remaining / 0 lost	12.9 (-0.1 points)
Epoch AI	169 → 167	Up 2	431 remaining / 13 lost	13.5 (+0.5 points)
Vals AI	173 → 166	Up 7	443 remaining / 1 lost	13.4 (+0.3 points)
Provider reports · Anthropic	173 → 173	No movement	444 remaining / 0 lost	13.0 (-0.0 points)
Provider reports · Moonshot AI	173 → 173	No movement	444 remaining / 0 lost	12.7 (-0.3 points)
Provider reports · Meta	173 → 173	No movement	444 remaining / 0 lost	13.0 (-0.0 points)
Provider reports · Mistral	173 → 173	No movement	444 remaining / 0 lost	13.0 (+0.0 points)
Provider reports · OpenAI	173 → 173	No movement	444 remaining / 0 lost	13.1 (+0.1 points)
All provider reports	173 → 173	No movement	444 remaining / 0 lost	12.8 (-0.2 points)

Lost support means a configuration no longer has a comparison path to the complete reference panel. Its rank is unavailable. A surviving path alone does not establish leaderboard eligibility.

Overlapping sources may remain correlated even with source balancing. Leave-one-source-out movement is a sensitivity diagnostic, not a spread of source-specific ranks or a confidence measure. Scoring methodology · Dated configuration audit

Every benchmark family

Removed evidence	Recomputed score	Change
deepswe	10.4	-2.6
arc-agi	14.5	+1.5
frontiercode	14.5	+1.5
mmmu-pro	13.9	+0.9
livebench-data	12.1	-0.9
hle	13.9	+0.8
mmlu-pro	13.5	+0.5
simpleqa	13.5	+0.5
livebench-instructions	13.5	+0.5
epoch-game-puzzles	12.7	-0.3
multi-swe-bench	12.7	-0.3
vibe-code	12.8	-0.2
livebench-reasoning	13.3	+0.2
livebench-math	13.2	+0.2
livebench-coding	13.2	+0.2
gpqa	12.8	-0.2
ifbench	12.9	-0.2
omniscience	12.9	-0.2
swe-atlas-qna	12.9	-0.1
vals-finance-agent	12.9	-0.1
enigma-eval	12.9	-0.1
aime	12.9	-0.1
vals-excel-modeling	12.9	-0.1
analyst-agent	13.1	+0.1
livebench-language	13.1	+0.1
automationbench	12.9	-0.1
harvey	12.9	-0.1
swe-rebench	12.9	-0.1
briefcase	13.1	+0.1
vals-legal-research	13.1	+0.1
apex-agents	12.9	-0.1
terminal-bench-science	13.1	+0.1
itbench	13.0	-0.1
swe-atlas-test-writing	13.0	-0.1
gdp-pdf	13.0	-0.1
vals-code-migration	13.0	-0.0
terminal-bench	13.0	-0.0
arc-agi-3	13.0	-0.0
enterprise-ops	13.1	+0.0
critpt	13.0	-0.0
gmmlu	13.0	+0.0
aa-lcr	13.0	+0.0
gdpval	13.0	+0.0
livecodebench	13.0	+0.0
frontiermath	13.0	-0.0
math500	13.0	+0.0
swe-atlas-refactoring	13.0	-0.0
vals-tax-agent	13.0	+0.0
swe-bench-pro	13.0	+0.0
tau-bench	13.0	-0.0

What evidence is missing from the fit?

Across all collected settings for this model: 95 observations, 86 contributing, 0 matched without graph weight and 9 excluded. Counts do not establish rank eligibility.

3 · Benchmark family or source protocol has not been reviewed
1 · Original SciCode grading defects are under review; retained as supporting evidence, excluded from aggregate and capability scores. Corrected SciCode-Verified requires a separate protocol review. Sources: https://arxiv.org/abs/2608.04975 and https://artificialanalysis.ai/evaluations/scicode
1 · Evaluator has not been reviewed for independent admission
1 · Effort is not identified
1 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately
2 · Supporting evidence outside the reviewed capability core

Inspect every observation and exclusion →

Benchmark results

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.

Benchmark / source	Low	Medium	High	XHigh	Non-reasoning	Unspecified
AA-AnalystAgentArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of tasks solved on all 5 of 5 attempts · 80 tasks, 5 attempts per task · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	15% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
AA-Briefcase EloArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Briefcase v1.1 is an agentic knowledge work benchmark developed by Artificial Analysis. AA-Briefcase Elo is a combined metric that aggregates rubric pass rate, analytical quality Elo and presentation Elo · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-10-18.	Not reported	Not reported	Not reported	719.36 Elo Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
APEX-Agents-AAArtificial 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 2026-10-18.	Not reported	Not reported	Not reported	28.17% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
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 2026-10-18.	Not reported	67% Reported settings GPT-5.4 mini (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-medium	Not reported	77% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	37% Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
AutomationBench-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of task objectives completed with no guardrail violations · Higher is better · Benchmark developed by Zapier · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	26.2% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
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. CritPt is under review by its authors following external feedback · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	2.86% Reported settings GPT-5.4 mini (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-medium	Not reported	10% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	0% Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
EnterpriseOps-Gym-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Overall task success rate across all eight enterprise domains (oracle tool mode) · Benchmark developed by ServiceNow Research · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	34.62% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
GDP.pdf: All-pass RateArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of attempts where every atomic criterion passed · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	13.4% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
GDPval-AA v2.1 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 DeepSeek V4.1 Flash (max) at 1600 · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-10-18.	Not reported	Not reported	Not reported	1,011.97 Elo Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	589.8 Elo Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
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 2026-10-18.	Not reported	Not reported	Not reported	89.41% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
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 2026-10-18.	Not reported	82.32% Reported settings GPT-5.4 mini (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-medium	Not reported	87.47% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	60.61% Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
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 2026-10-18.	Not reported	18.58% Reported settings GPT-5.4 mini (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-medium	Not reported	28.13% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	5.89% Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
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 2026-10-18.	Not reported	64.76% Reported settings GPT-5.4 mini (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-medium	Not reported	73.27% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	38.84% Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
ITBench-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Combined public and private Site Reliability Engineering (SRE) tasks · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	35.22% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	18.64% Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
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 2026-10-18.	Not reported	Not reported	Not reported	3.89% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
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 2026-10-18.	Not reported	71.16% Reported settings GPT-5.4 mini (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-medium	Not reported	73.29% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	60.46% Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	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 2026-10-18.	Not reported	-20.62 Reported settings GPT-5.4 mini (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-medium	Not reported	-18.92 Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	-45.42 Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
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 2026-10-18.	Not reported	36.47% Reported settings GPT-5.4 mini (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-medium	Not reported	37.47% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	25.65% Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
SciCodeArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark curated by scientists across 16 disciplines. SciCode is under review following an independent audit of the dataset · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	52.08% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
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 2026-10-18.	Not reported	36.55% Reported settings GPT-5.4 mini (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-medium	Not reported	83.33% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	23.39% Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
𝜏³-BankingArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by Sierra Research · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	25.57% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
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 2026-10-18.	Not reported	34.09% Reported settings GPT-5.4 mini (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-medium	Not reported	52.27% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	18.18% Reported settings GPT-5.4 mini (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini-non-reasoning	Not reported
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 2026-10-18.	Not reported	Not reported	Not reported	59.18% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
Terminal-Bench v4.0Artificial Analysis · mini-SWE-agent v2.4.6; 66 tasks; pass@1 averaged over 3 repeats; 500 steps; task timeout up to 8 hours Benchmark details Benchmark developed by the Laude Institute, Stanford, and open-source contributors · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	2.02% Reported settings GPT-5.4 mini (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-4-mini	Not reported	Not reported
ARC-AGI-2ARC Prize · v2 Semi-Private Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-10-18.	1.11% Reported settings GPT-5.4 Mini (Low); Source model ID: gpt-5-4-mini-low; 120 tasks	4.44% Reported settings GPT-5.4 Mini (Medium); Source model ID: gpt-5-4-mini-medium; 120 tasks	13.19% Reported settings GPT-5.4 Mini (High); Source model ID: gpt-5-4-mini-high; 120 tasks	18.9% Reported settings GPT-5.4 Mini (XHigh); Source model ID: gpt-5-4-mini-xhigh; 120 tasks	Not reported	Not reported

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 · 8,915 results across 259 models · Collected 2026-10-18

GPT-5.4 Mini · XHigh

Compare with

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context

Capability score evidence

Benchmark results