RankingEffort benchmarks

GPT-6 Sol

Compare models

147 measured results · 7 effort settings · 64 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-6 Sol · Max4 of 7 capabilities supported

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

Agentic
A55.1Supported
Hard reasoning
A68.4Supported
Coding
A74.5Supported
Human pref
Unknown
Knowledge
A63.7Supported
Multimodal
A56.0Preliminary
Long context
A79.6Preliminary

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-6 Sol · Max: 66.4 overall comparison score. Meets the overall evidence requirements.

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

Shared-result disagreement. This aggregate order conflicts with matched benchmark results against 2 other configurations. Different test coverage and opponents affect the fit. Inspect published results before treating the rank as conclusive.

Which sources carry the weight?

OpenAI · 23.6% direct comparison weight · provider report
Epoch AI · 11.9% direct comparison weight · independent of the model provider
Cognition · 11.7% direct comparison weight · independent of the model provider
LiveBench · 11.7% direct comparison weight · independent of the model provider
Vals AI · 11.4% direct comparison weight · independent of the model provider
ARC Prize · 11.3% direct comparison weight · independent of the model provider
Artificial Analysis · 10.0% direct comparison weight · independent of the model provider
Terminal-Bench · 8.4% direct comparison weight · independent of the model provider

Which benchmark families carry the weight?

FrontierCode · 11.7% direct comparison weight
Terminal-Bench · 10.5% direct comparison weight
ARC-AGI · 5.9% direct comparison weight
AutomationBench · 5.5% direct comparison weight
GDP.pdf · 5.5% direct comparison weight
Other families (30) · 60.9% 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 2026-10-31.

Both ranks use the same surviving configurations for each removal. The cohort starts with 446 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	43 → 43	No movement	203 remaining / 243 lost	67.3 (+0.9 points)
ARC Prize	43 → 47	Down 4	446 remaining / 0 lost	68.5 (+2.1 points)
Datacurve	43 → 42	Up 1	446 remaining / 0 lost	62.1 (-4.3 points)
LiveCodeBench	43 → 43	No movement	446 remaining / 0 lost	66.4 (-0.0 points)
Scale	43 → 43	No movement	446 remaining / 0 lost	67.4 (+1.0 points)
Terminal-Bench	43 → 46	Down 3	446 remaining / 0 lost	64.3 (-2.1 points)
Cognition	43 → 42	Up 1	446 remaining / 0 lost	62.1 (-4.3 points)
LiveBench	43 → 38	Up 5	445 remaining / 1 lost	70.5 (+4.1 points)
SWE-rebench	43 → 43	No movement	446 remaining / 0 lost	65.9 (-0.5 points)
Epoch AI	43 → 45	Down 2	433 remaining / 13 lost	65.8 (-0.6 points)
Vals AI	43 → 38	Up 5	445 remaining / 1 lost	70.7 (+4.3 points)
Provider reports · Anthropic	43 → 43	No movement	446 remaining / 0 lost	66.3 (-0.1 points)
Provider reports · Moonshot AI	43 → 43	No movement	446 remaining / 0 lost	66.1 (-0.3 points)
Provider reports · Meta	43 → 42	Up 1	446 remaining / 0 lost	66.3 (-0.1 points)
Provider reports · Mistral	43 → 43	No movement	446 remaining / 0 lost	66.4 (+0.0 points)
Provider reports · OpenAI	43 → 36	Up 7	446 remaining / 0 lost	72.6 (+6.2 points)
All provider reports	43 → 35	Up 8	446 remaining / 0 lost	72.1 (+5.7 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
frontiercode	62.1	-4.3
deepswe	62.1	-4.3
terminal-bench	63.5	-3.0
arc-agi	68.4	+2.0
vals-finance-agent	67.7	+1.3
vals-legal-research	67.6	+1.2
vibe-code	65.3	-1.1
vals-tax-agent	67.6	+1.1
vals-code-migration	65.4	-1.1
simpleqa	67.5	+1.0
livebench-instructions	67.2	+0.8
vals-excel-modeling	65.6	-0.8
mmmu-pro	67.2	+0.7
frontiermath	65.7	-0.7
hle	67.0	+0.6
arc-agi-3	67.0	+0.6
swe-rebench	65.9	-0.5
livebench-data	65.9	-0.5
livebench-coding	66.0	-0.4
epoch-game-puzzles	66.1	-0.3
multi-swe-bench	66.7	+0.3
swe-atlas-qna	66.2	-0.3
mmlu-pro	66.7	+0.2
ifbench	66.2	-0.2
swe-atlas-test-writing	66.3	-0.2
automationbench	66.3	-0.2
tau-bench	66.3	-0.2
omniscience	66.6	+0.2
aime	66.3	-0.1
terminal-bench-science	66.3	-0.1
livebench-math	66.3	-0.1
gdpval	66.5	+0.1
livebench-language	66.3	-0.1
critpt	66.5	+0.1
itbench	66.4	-0.1
gdp-pdf	66.5	+0.1
aa-lcr	66.5	+0.1
gpqa	66.4	-0.1
analyst-agent	66.5	+0.1
enterprise-ops	66.5	+0.1
apex-agents	66.5	+0.0
swe-atlas-refactoring	66.4	-0.0
enigma-eval	66.4	-0.0
briefcase	66.4	-0.0
harvey	66.4	+0.0
livebench-reasoning	66.4	+0.0
livecodebench	66.4	-0.0
math500	66.4	-0.0
swe-bench-pro	66.4	+0.0
gmmlu	66.4	-0.0

What evidence is missing from the fit?

Across all collected settings for this model: 221 observations, 158 contributing, 0 matched without graph weight and 63 excluded. Counts do not establish rank eligibility.

6 · 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
4 · Benchmark family or source protocol has not been reviewed
4 · Evaluator has not been reviewed for independent admission
25 · Provider-published lab self-report. Shown as a labeled claim and not admitted as a matched board comparison.
5 · Private user-flagged prompt protocol needs separate review before capability scoring; retained as an exact lower-is-better sourced claim.
4 · Safety failure behavior is not a capability or intelligence score.
13 · Supporting evidence outside the reviewed capability core
2 · 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.

Reviewed retained observations are in use. October 2026 Real-SWE board has no 95% confidence whisker, so these scores stay at the 2026-10-18 retrieval. Original source retrieval 2026-10-18; source updated 2026-10-18.

Benchmark / source	Low	Medium	High	XHigh	Max	None	Non-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 2026-10-31.	885.28 Elo Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	1,147.76 Elo Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	1,266.21 Elo Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	1,358.03 Elo Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	1,475.81 Elo Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	1,104.61 Elo Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
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 2026-10-31.	79.33% Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	82.33% Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	83.67% Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	81.33% Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	83.67% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	64% Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
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 2026-10-31.	53.89% Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	57.97% Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	60.13% Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	61.69% Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	61.63% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	34.18% Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
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 2026-10-31.	16.29% Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	24.57% Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	25.43% Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	28% Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	30.86% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	4% Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
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 2026-10-31.	24.2% Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	23.8% Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	24.4% Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	24.6% Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	25.2% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	17% Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
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 2026-10-31.	1,201.12 Elo Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	1,351.87 Elo Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	1,403.81 Elo Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	1,466.31 Elo Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	1,515.88 Elo Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	1,254.28 Elo Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
Harvey LAB-AA v1.1: Hallucination-Gated All-Pass RateArtificial Analysis · Stirrup; Harvey private dataset v1.1.0; 120 tasks across 24 practice areas; up to 200 turns; internet disabled; GPT-6 Sol, Grok 4.7 and Claude Opus 5.5 rubric judges; GPT-6 Sol High two-pass hallucination check; hallucination-gated all-pass averaged across judges Benchmark details Average over tasks of the share of the judge panel finding every rubric criterion passed, with a task zeroed on any material hallucination · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-31.	Not reported	Not reported	Not reported	Not reported	3.61% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	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 2026-10-31.	34.94% Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	40.96% Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	44.11% Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	46.29% Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	47.91% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	18.4% Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
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 2026-10-31.	Not reported	Not reported	Not reported	Not reported	49.44% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	Not reported
MLCR-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of responses judged complete with no claim contradicting the record · Expert + compound (hard) sets · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-31.	25.56% Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	28.89% Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	29.44% Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	32.22% Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	33.89% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	22.78% Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
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 2026-10-31.	80.17% Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	80.4% Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	81.97% Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	82.54% Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	82.95% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	70.35% Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
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 2026-10-31.	50.23% Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	53.82% Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	54.86% Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	55.09% Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	57.64% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	47.34% Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
Terminal-Bench-Science 0.1Artificial Analysis · mini-swe-agent; Terminal-Bench-Science 0.1.0; 70 tasks; pass@1 averaged over 3 repeats Benchmark details Share of tasks passed (pass@1), averaged over three repeats · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-31.	3.33% Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	8.57% Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	17.14% Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	17.62% Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	30% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	5.71% Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
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 2026-10-31.	9.09% Reported settings GPT-6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-6-sol-low	18.69% Reported settings GPT-6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-6-sol-medium	26.26% Reported settings GPT-6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-6-sol-high	30.3% Reported settings GPT-6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-6-sol-xhigh	43.94% Reported settings GPT-6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-6-sol	Not reported	13.13% Reported settings GPT-6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-6-sol-non-reasoning
ARC-AGI-2ARC Prize · v2 Semi-Private Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-10-31.	31.53% Reported settings GPT-6 Sol (Low); Source model ID: openai-gpt-6-sol-low; 120 tasks	57.78% Reported settings GPT-6 Sol (Medium); Source model ID: openai-gpt-6-sol-medium; 120 tasks	68.89% Reported settings GPT-6 Sol (High); Source model ID: openai-gpt-6-sol-high; 120 tasks	78.06% Reported settings GPT-6 Sol (XHigh); Source model ID: openai-gpt-6-sol-xhigh; 120 tasks	89.58% Reported settings GPT-6 Sol (Max); Source model ID: openai-gpt-6-sol-max; 120 tasks	1.67% Reported settings GPT-6 Sol (None); Source model ID: openai-gpt-6-sol-none; 120 tasks	Not reported
ARC-AGI-3 · provider-adapterARC Prize · Provider Adapter Benchmark details ARC-AGI-3; Provider Adapter. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-31.	1.76% Reported settings ARC-AGI-3 Semi-Private verified summary; Provider Adapter harness; source group openai-gpt-6-sol; variant provider-adapter-low; public-demo task scores are excluded	4.93% Reported settings ARC-AGI-3 Semi-Private verified summary; Provider Adapter harness; source group openai-gpt-6-sol; variant provider-adapter-medium; public-demo task scores are excluded	5.86% Reported settings ARC-AGI-3 Semi-Private verified summary; Provider Adapter harness; source group openai-gpt-6-sol; variant provider-adapter-high; public-demo task scores are excluded	9.45% Reported settings ARC-AGI-3 Semi-Private verified summary; Provider Adapter harness; source group openai-gpt-6-sol; variant provider-adapter-xhigh; public-demo task scores are excluded	23.03% Reported settings ARC-AGI-3 Semi-Private verified summary; Provider Adapter harness; source group openai-gpt-6-sol; variant provider-adapter-max; public-demo task scores are excluded	1.21% Reported settings ARC-AGI-3 Semi-Private verified summary; Provider Adapter harness; source group openai-gpt-6-sol; variant provider-adapter-none; public-demo task scores are excluded	Not reported
ARC-AGI-3 · standardARC Prize · Standard Benchmark details ARC-AGI-3; Standard. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-31.	0.13% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-6-sol; variant low; public-demo task scores are excluded	0.42% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-6-sol; variant medium; public-demo task scores are excluded	0.9% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-6-sol; variant high; public-demo task scores are excluded	1.8% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-6-sol; variant xhigh; public-demo task scores are excluded	4.62% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-6-sol; variant max; public-demo task scores are excluded	0.28% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-6-sol; variant none; public-demo task scores are excluded	Not reported
FrontierCode 1.1 · extended-codexCognition · codex Benchmark details FrontierCode 1.1; codex. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-31.	50.55% Reported settings FrontierCode 1.1; extended; 150 tasks; codex; source effort low; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0.0013	57.11% Reported settings FrontierCode 1.1; extended; 150 tasks; codex; source effort medium; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0.0013	59.62% Reported settings FrontierCode 1.1; extended; 150 tasks; codex; source effort high; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	59.13% Reported settings FrontierCode 1.1; extended; 150 tasks; codex; source effort xhigh; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0.0013	60.69% Reported settings FrontierCode 1.1; extended; 150 tasks; codex; source effort max; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	Not reported	Not reported
FrontierCode 1.1 · main-codexCognition · codex Benchmark details FrontierCode 1.1; codex. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-31.	37.28% Reported settings FrontierCode 1.1; main; 100 tasks; codex; source effort low; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0.002	45.92% Reported settings FrontierCode 1.1; main; 100 tasks; codex; source effort medium; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	47.7% Reported settings FrontierCode 1.1; main; 100 tasks; codex; source effort high; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	48.45% Reported settings FrontierCode 1.1; main; 100 tasks; codex; source effort xhigh; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0.002	49.27% Reported settings FrontierCode 1.1; main; 100 tasks; codex; source effort max; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	Not reported	Not reported
Terminal-Bench 4.0 · codexTerminal-Bench · Codex Benchmark details Terminal-Bench 4.0; Codex. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-31.	Not reported	Not reported	Not reported	Not reported	49.4% Reported settings Terminal-Bench 4.0.0; 66 tasks; Codex; effort max; reported margin ±3.2 percentage points, confidence level unspecified; release date 2026-09-22; board checked 2026-10-23	Not reported	Not reported
LiveBench · AMPS Hard · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-31.	Not reported	Not reported	Not reported	Not reported	98% Reported settings LiveBench official task evaluation; livebench-2026-06-25-AMPS_Hard; gpt-6-sol-max; checkpoint=gpt-6-sol	Not reported	Not reported
LiveBench · code completion · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-31.	Not reported	Not reported	Not reported	Not reported	80.44% Reported settings LiveBench official task evaluation; livebench-2026-06-25-code_completion; gpt-6-sol-max; checkpoint=gpt-6-sol	Not reported	Not reported
LiveBench · code generation · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-31.	Not reported	Not reported	Not reported	Not reported	83.1% Reported settings LiveBench official task evaluation; livebench-2026-06-25-code_generation; gpt-6-sol-max; checkpoint=gpt-6-sol	Not reported	Not reported
LiveBench · connections · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-31.	Not reported	Not reported	Not reported	Not reported	100% Reported settings LiveBench official task evaluation; livebench-2026-06-25-connections; gpt-6-sol-max; checkpoint=gpt-6-sol	Not reported	Not reported
LiveBench · consecutive events · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-31.	Not reported	Not reported	Not reported	Not reported	89.55% Reported settings LiveBench official task evaluation; livebench-2026-06-25-consecutive_events; gpt-6-sol-max; checkpoint=gpt-6-sol	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 · 8,399 results across 259 models · Collected 2026-10-31

GPT-6 Sol

Compare with

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context

Capability score evidence

Benchmark results