RankingEffort benchmarks

GPT-5.6 Sol · XHigh

Compare models

205 measured results · 8 effort settings · 85 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-5.6 Sol · XHigh4 of 7 capabilities supported

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

Agentic
A55.8Supported
Hard reasoning
A72.9Supported
Coding
A78.2Supported
Human pref
Unknown
Knowledge
A42.1Supported
Multimodal
A55.1Preliminary
Long context
A61.3Preliminary

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-5.6 Sol · XHigh: 79.3 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 1 other configurations. Different test coverage and opponents affect the fit. Inspect published results before treating the rank as conclusive.

Which sources carry the weight?

Cognition · 27.6% direct comparison weight · independent of the model provider
ARC Prize · 26.3% direct comparison weight · independent of the model provider
Artificial Analysis · 24.8% direct comparison weight · independent of the model provider
Datacurve · 15.4% direct comparison weight · independent of the model provider
Scale · 5.9% direct comparison weight · independent of the model provider

Which benchmark families carry the weight?

FrontierCode · 27.6% direct comparison weight
ARC-AGI · 26.3% direct comparison weight
DeepSWE · 15.4% direct comparison weight
SWE-Atlas Codebase QnA · 5.9% direct comparison weight
IFBench · 2.2% direct comparison weight
Other families (12) · 22.5% 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.

DeepSWE v1.1 review: Epoch AI reported grading errors in at least 23 of 113 tasks. These published results remain included. Review their limitations in Epoch's review when interpreting this evidence.

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	22 → 18	Up 4	203 remaining / 241 lost	83.3 (+4.0 points)
ARC Prize	22 → 26	Down 4	444 remaining / 0 lost	79.2 (-0.1 points)
Datacurve	22 → 24	Down 2	444 remaining / 0 lost	75.2 (-4.1 points)
LiveCodeBench	22 → 23	Down 1	444 remaining / 0 lost	79.3 (-0.0 points)
Scale	22 → 25	Down 3	444 remaining / 0 lost	78.5 (-0.8 points)
Terminal-Bench	22 → 20	Up 2	444 remaining / 0 lost	79.3 (+0.0 points)
Cognition	22 → 20	Up 2	444 remaining / 0 lost	79.8 (+0.5 points)
LiveBench	22 → 25	Down 3	443 remaining / 1 lost	80.0 (+0.7 points)
SWE-rebench	22 → 22	No movement	444 remaining / 0 lost	79.0 (-0.2 points)
Epoch AI	22 → 24	Down 2	431 remaining / 13 lost	79.8 (+0.5 points)
Vals AI	22 → 19	Up 3	443 remaining / 1 lost	79.3 (+0.1 points)
Provider reports · Anthropic	22 → 23	Down 1	444 remaining / 0 lost	79.3 (+0.0 points)
Provider reports · Moonshot AI	22 → 21	Up 1	444 remaining / 0 lost	79.5 (+0.2 points)
Provider reports · Meta	22 → 22	No movement	444 remaining / 0 lost	79.3 (-0.0 points)
Provider reports · Mistral	22 → 22	No movement	444 remaining / 0 lost	79.3 (+0.0 points)
Provider reports · OpenAI	22 → 22	No movement	444 remaining / 0 lost	79.0 (-0.3 points)
All provider reports	22 → 22	No movement	444 remaining / 0 lost	79.1 (-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
deepswe	75.1	-4.1
swe-atlas-qna	77.4	-1.9
frontiercode	79.8	+0.5
simpleqa	79.8	+0.5
swe-atlas-refactoring	79.8	+0.5
mmmu-pro	79.7	+0.5
swe-atlas-test-writing	79.7	+0.4
tau-bench	79.6	+0.3
ifbench	79.6	+0.3
arc-agi-3	79.0	-0.3
gpqa	79.0	-0.2
epoch-game-puzzles	79.0	-0.2
swe-rebench	79.0	-0.2
gdp-pdf	79.0	-0.2
hle	79.5	+0.2
livebench-instructions	79.5	+0.2
mmlu-pro	79.5	+0.2
critpt	79.4	+0.2
terminal-bench	79.4	+0.2
multi-swe-bench	79.2	-0.1
aime	79.2	-0.1
swe-bench-pro	79.4	+0.1
vibe-code	79.2	-0.1
gdpval	79.4	+0.1
enigma-eval	79.4	+0.1
aa-lcr	79.4	+0.1
vals-code-migration	79.2	-0.1
livebench-coding	79.2	-0.1
automationbench	79.4	+0.1
livebench-language	79.3	+0.1
livebench-reasoning	79.2	-0.1
briefcase	79.3	+0.1
omniscience	79.2	-0.1
harvey	79.2	-0.1
vals-excel-modeling	79.2	-0.1
frontiermath	79.2	-0.1
livecodebench	79.2	-0.0
vals-finance-agent	79.3	+0.0
terminal-bench-science	79.2	-0.0
enterprise-ops	79.2	-0.0
arc-agi	79.2	-0.0
itbench	79.2	-0.0
livebench-data	79.3	-0.0
apex-agents	79.3	+0.0
gmmlu	79.3	-0.0
math500	79.3	-0.0
analyst-agent	79.3	-0.0
livebench-math	79.3	+0.0
vals-legal-research	79.3	-0.0
vals-tax-agent	79.3	+0.0

What evidence is missing from the fit?

Across all collected settings for this model: 474 observations, 187 contributing, 0 matched without graph weight and 287 excluded. Counts do not establish rank eligibility.

11 · Benchmark family or source protocol has not been reviewed
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
11 · Evaluator has not been reviewed for independent admission
1 · Starred protocol annotation requires row-specific review before ranking.
3 · Effort is not identified
57 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately
43 · Narrow specialist biological, medical or security evaluation; preserved as raw evidence outside general capability scoring.
2 · Historical provider claim; secondary and superseded for current FrontierSWE comparisons by the independent Proximal FrontierSWE v2 mean@5 board at https://www.frontierswe.com/. Version-unspecified claims are not v2 results. Excluded from weighted Overall.
1 · System card section 8.6 cites the public Codex CLI result for Sol, while the Claude rows are internal Claude Code --bare reruns. Different agent harnesses and runs cannot form a matched comparison.
82 · Supporting evidence outside the reviewed capability core
1 · Provider-sourced OSWorld task revision is unverified; cannot join a known-version comparison.
7 · Kimi K3 Evaluation Details cites Artificial Analysis; retain as published context, not a new Moonshot comparison.
1 · Kimi K3 Evaluation Details cites GLM release blog or Vals AI, per model; retain as published context, not a new Moonshot comparison.
1 · Kimi K3 Evaluation Details cites GLM release blog, Artificial Analysis or OpenAI, per model; retain as published context, not a new Moonshot comparison.
1 · Kimi K3 Evaluation Details cites Anthropic or OpenAI release reports; retain as published context, not a new Moonshot comparison.
1 · Kimi K3 Evaluation Details cites official Agents Last Exam leaderboard; retain as published context, not a new Moonshot comparison.
1 · Kimi K3 Evaluation Details cites Artificial Analysis / APEX-Agents leaderboard; retain as published context, not a new Moonshot comparison.
3 · Kimi K3 Evaluation Details cites Vals AI; retain as published context, not a new Moonshot comparison.
12 · Composite index or publisher-specific internal evaluation lacks a stable general-capability comparison unit.
1 · Different native coding harnesses are mixed; exact shared agent configuration is not established.
23 · No matched, identified effort unit; duplicates and unnamed comparators do not add evidence
2 · No matched opponent in this evaluation unit
2 · The Qwen card cites an external leaderboard or release report for this comparator; it is not a new Qwen evaluation.
2 · Composite index or publisher-specific internal evaluation lacks a stable general-capability comparison unit. Composite index excluded to avoid double counting its ingredients.
5 · The launch table does not establish matching agent harnesses and evaluation settings. Retained as a provider claim; not a new comparable evaluation.
1 · Secondary Artificial Analysis leaderboard report; not an additional independent evaluation. Use the original evaluator for reviewed exact-configuration comparisons.
1 · Secondary Harvey leaderboard report; not an additional independent evaluation. Use the original Vals AI evaluation with its reviewed effort settings.
1 · The GLM card attributes this evaluation to Artificial Analysis; it is not a new Z.ai comparison.
4 · Legacy self-report lacks a reviewed comparison configuration

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	Max	None	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	Not reported	47.5% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	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.	1,043.52 Elo Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	1,241.93 Elo Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	1,368.7 Elo Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	1,437.45 Elo Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	1,480.04 Elo Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	1,008.14 Elo Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-non-reasoning	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.	78% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	80.33% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	81.67% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	82.33% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	84% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	62.33% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-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.	40.99% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	51.33% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	55.31% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	55.31% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	60.08% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	22.66% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-non-reasoning	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.	14.86% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	22.86% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	25.71% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	28.57% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	32.29% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	5.14% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-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	Not reported	42.91% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	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.	21% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	26.2% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	27.8% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	27.6% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	27.2% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	15.2% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-non-reasoning	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.	1,303.39 Elo Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	1,421.47 Elo Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	1,503.99 Elo Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	1,568.48 Elo Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	1,608.63 Elo Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	1,241.59 Elo Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-non-reasoning	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.	89.8% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	92.63% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	92.83% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	93.13% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	94.14% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	78.99% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-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.	39.39% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	42.22% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	46.01% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	47.31% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	49.49% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	16.68% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-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.	66.53% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	69.59% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	69.18% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	71.02% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	72.65% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	Not reported	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	Not reported	56.21% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	Not reported	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.	12.78% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	15% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	17.22% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	18.33% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	26.11% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	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.	80.98% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	81.39% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	81.85% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	82.66% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	83.41% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	71.91% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-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.	18.87 Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	19.4 Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	20.37 Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	20.98 Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	21.97 Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	1.1 Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-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.	57.17% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	57.75% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	58.35% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	58.82% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	59.4% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	48.7% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-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.	56.37% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	57.41% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	57.75% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	57.06% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	57.06% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	47.69% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-non-reasoning	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.	76.02% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	80.99% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	83.33% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	84.8% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	85.09% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	Not reported	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.	29.07% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	36.49% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	36.7% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	38.14% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	44.33% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	19.59% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-non-reasoning	Not reported
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 2026-10-18.	Not reported	Not reported	Not reported	Not reported	22.38% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	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.	60.61% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	62.88% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	62.12% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	61.36% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	65.91% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	Not reported	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.	76.78% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	86.14% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	87.27% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	89.51% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	88.01% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	74.16% Reported settings GPT-5.6 Sol (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-non-reasoning	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.	1.01% Reported settings GPT-5.6 Sol (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-low	14.65% Reported settings GPT-5.6 Sol (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-medium	20.71% Reported settings GPT-5.6 Sol (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-high	24.75% Reported settings GPT-5.6 Sol (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol-xhigh	39.9% Reported settings GPT-5.6 Sol (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-sol	Not reported	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.	42.5% Reported settings GPT-5.6 Sol (Low); Source model ID: openai-gpt-5-6-sol-low; 120 tasks	67.08% Reported settings GPT-5.6 Sol (Medium); Source model ID: openai-gpt-5-6-sol-medium; 120 tasks	85.42% Reported settings GPT-5.6 Sol (High); Source model ID: openai-gpt-5-6-sol-high; 120 tasks	90% Reported settings GPT-5.6 Sol (XHigh); Source model ID: openai-gpt-5-6-sol-xhigh; 120 tasks	92.5% Reported settings GPT-5.6 Sol (Max); Source model ID: openai-gpt-5-6-sol-max; 120 tasks	Not reported	Not reported	Not reported
DeepSWE v1.1Datacurve · mini-swe-agent Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-10-18.	45.4% Reported settings gpt-5-6-sol; Reported effort: low; 113 unique tasks; 452 attempts; 4 runs	61.1% Reported settings gpt-5-6-sol; Reported effort: medium; 113 unique tasks; 452 attempts; 4 runs	69.4% Reported settings gpt-5-6-sol; Reported effort: high; 113 unique tasks; 451 attempts; 4 runs	70.7% Reported settings gpt-5-6-sol; Reported effort: xhigh; 113 unique tasks; 451 attempts; 4 runs	72.7% Reported settings gpt-5-6-sol; Reported effort: max; 113 unique tasks; 450 attempts; 4 runs	Not reported	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-5.6 Sol · XHigh

Compare with

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context

Capability score evidence

Benchmark results