RankingEffort benchmarks

o3-mini

Compare models

32 measured results · 4 effort settings · 22 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

Ao3-mini · High4 of 7 capabilities supported

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

Agentic
A4.7Supported
Hard reasoning
A9.0Supported
Coding
A5.5Supported
Human pref
Unknown
Knowledge
A10.9Supported
Multimodal
Unknown
Long context
A7.5Preliminary

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

o3-mini · High: 2.5 overall comparison score. Meets the overall evidence requirements.

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

Which sources carry the weight?

Artificial Analysis · 43.2% direct comparison weight · independent of the model provider
ARC Prize · 36.8% direct comparison weight · independent of the model provider
Scale · 20.0% direct comparison weight · independent of the model provider

Which benchmark families carry the weight?

ARC-AGI · 36.8% direct comparison weight
Humanity’s Last Exam · 22.9% direct comparison weight
MMLU-Pro · 2.9% direct comparison weight
LiveCodeBench · 2.9% direct comparison weight
math500 · 2.9% direct comparison weight
Other families (11) · 31.5% 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 2026-10-23.

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	182 → 179	Up 3	203 remaining / 241 lost	2.1 (-0.4 points)
ARC Prize	294 → 306	Down 12	444 remaining / 0 lost	2.6 (+0.1 points)
Datacurve	294 → 294	No movement	444 remaining / 0 lost	1.9 (-0.5 points)
LiveCodeBench	294 → 294	No movement	444 remaining / 0 lost	2.5 (+0.0 points)
Scale	294 → 281	Up 13	444 remaining / 0 lost	3.6 (+1.1 points)
Terminal-Bench	294 → 294	No movement	444 remaining / 0 lost	2.4 (-0.0 points)
Cognition	294 → 295	Down 1	444 remaining / 0 lost	2.4 (-0.1 points)
LiveBench	293 → 293	No movement	443 remaining / 1 lost	2.6 (+0.1 points)
SWE-rebench	294 → 296	Down 2	444 remaining / 0 lost	2.5 (+0.0 points)
Epoch AI	284 → 287	Down 3	431 remaining / 13 lost	2.5 (-0.0 points)
Vals AI	293 → 294	Down 1	443 remaining / 1 lost	2.3 (-0.1 points)
Provider reports · Anthropic	294 → 294	No movement	444 remaining / 0 lost	2.5 (-0.0 points)
Provider reports · Moonshot AI	294 → 294	No movement	444 remaining / 0 lost	2.4 (-0.1 points)
Provider reports · Meta	294 → 294	No movement	444 remaining / 0 lost	2.5 (-0.0 points)
Provider reports · Mistral	294 → 294	No movement	444 remaining / 0 lost	2.5 (+0.0 points)
Provider reports · OpenAI	294 → 294	No movement	444 remaining / 0 lost	2.5 (+0.0 points)
All provider reports	294 → 294	No movement	444 remaining / 0 lost	2.4 (-0.1 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
hle	3.3	+0.9
deepswe	1.9	-0.5
omniscience	2.2	-0.2
mmmu-pro	2.7	+0.2
epoch-game-puzzles	2.3	-0.2
ifbench	2.3	-0.2
mmlu-pro	2.6	+0.1
simpleqa	2.6	+0.1
math500	2.4	-0.1
arc-agi-3	2.4	-0.1
enigma-eval	2.4	-0.1
gdpval	2.5	+0.1
gpqa	2.4	-0.1
frontiercode	2.4	-0.1
tau-bench	2.5	+0.1
terminal-bench	2.5	+0.1
livebench-instructions	2.5	+0.1
critpt	2.4	-0.1
gdp-pdf	2.5	+0.1
automationbench	2.5	+0.1
multi-swe-bench	2.4	-0.0
livebench-language	2.5	+0.0
swe-rebench	2.5	+0.0
livecodebench	2.4	-0.0
livebench-coding	2.4	-0.0
apex-agents	2.5	+0.0
harvey	2.4	-0.0
terminal-bench-science	2.5	+0.0
swe-atlas-qna	2.4	-0.0
briefcase	2.5	+0.0
analyst-agent	2.5	+0.0
vibe-code	2.4	-0.0
livebench-reasoning	2.4	-0.0
itbench	2.4	-0.0
aime	2.4	-0.0
aa-lcr	2.5	+0.0
vals-code-migration	2.5	-0.0
swe-atlas-test-writing	2.5	-0.0
vals-finance-agent	2.5	-0.0
frontiermath	2.5	+0.0
vals-legal-research	2.5	-0.0
arc-agi	2.5	+0.0
vals-tax-agent	2.5	-0.0
vals-excel-modeling	2.5	-0.0
gmmlu	2.5	-0.0
livebench-math	2.5	+0.0
swe-atlas-refactoring	2.5	+0.0
swe-bench-pro	2.5	+0.0
livebench-data	2.5	-0.0
enterprise-ops	2.5	+0.0

What evidence is missing from the fit?

Across all collected settings for this model: 35 observations, 30 contributing, 0 matched without graph weight and 5 excluded. Counts do not establish rank eligibility.

1 · 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
3 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately

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	Reasoning
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	141.71 Elo Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	Not reported	43% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	0.53% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	Not reported	0.29% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	2% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	230.33 Elo Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	Not reported	77.27% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	74.85% Reported settings o3-mini; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/o3-mini; Mode from evaluator isReasoning metadata; effort budget not specified
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	Not reported	12.04% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	7.91% Reported settings o3-mini; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/o3-mini; Mode from evaluator isReasoning metadata; effort budget not specified
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	Not reported	67.14% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	Not reported
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 2026-10-18.	Not reported	Not reported	73.44% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	71.75% Reported settings o3-mini; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/o3-mini; Mode from evaluator isReasoning metadata; effort budget not specified
MATH-500Artificial 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	98.47% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	97.33% Reported settings o3-mini; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/o3-mini; Mode from evaluator isReasoning metadata; effort budget not specified
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 2026-10-18.	Not reported	Not reported	80.19% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	79.06% Reported settings o3-mini; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/o3-mini; Mode from evaluator isReasoning metadata; effort budget not specified
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	Not reported	-42.57 Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	Not reported	21.27% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	42.82% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	Not reported	31.29% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	28.65% Reported settings o3-mini; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/o3-mini; Mode from evaluator isReasoning metadata; effort budget not specified
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	5.15% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	Not reported	6.06% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	6.82% Reported settings o3-mini; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/o3-mini; Mode from evaluator isReasoning metadata; effort budget not specified
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	4.49% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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	0% Reported settings o3-mini (High); Model source: https://artificialanalysis.ai/models/o3-mini-high	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.	0% Reported settings o3-mini (Low); Source model ID: o3-mini-2025-01-31-low; 120 tasks	2.08% Reported settings o3-mini (Medium); Source model ID: o3-mini-2025-01-31-medium; 120 tasks	2.99% Reported settings o3-mini (High); Source model ID: o3-mini-2025-01-31-high; 120 tasks	Not reported
HLE · text only · Scale official evaluationScale · Scale official evaluation Benchmark details Scale text-only public HLE subset; shares the full HLE family budget; automatic answer judging. Unit: percent. Source collected 2026-10-18.	Not reported	10.31% Reported settings Scale text-only public HLE subset; shares the full HLE family budget; automatic answer judging.; published row 2025-04-10T19:29:02.000Z; reported margin ±1.28 percentage points; Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.	13.37% Reported settings Scale text-only public HLE subset; shares the full HLE family budget; automatic answer judging.; published row 2025-04-10T19:29:02.000Z; reported margin ±1.44 percentage points; Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.	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 · 8,916 results across 259 models · Collected 2026-10-18

o3-mini

Compare with

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context

Capability score evidence

Benchmark results