RankingEffort benchmarks

GPT-5.6 Terra effort benchmarks

Compare measured Low, Medium, High and other settings. Every number keeps its source.

Model

GPT-5.6 Terra

182 measured results · 7 reported effort labels · 68 benchmark/harness combinations. Model evidence and sources →

Capability score evidence

GPT-5.6 Terra · Max: 54.7 overall comparison score. Meets the overall evidence requirements.

Which sources carry the weight?

LiveBench · 13.6% direct comparison weight · independent
Cognition · 13.5% direct comparison weight · independent
Vals AI · 13.4% direct comparison weight · independent
Terminal-Bench · 13.2% direct comparison weight · independent
Artificial Analysis · 12.8% direct comparison weight · independent
Epoch AI · 12.5% direct comparison weight · independent
ARC Prize · 12.3% direct comparison weight · independent
Datacurve · 8.7% direct comparison weight · independent

These shares describe the comparisons entering the fit, not fractions of the final score.

How sensitive is this score?

Remove one evaluator or family and refit the graph against the same overall panel. Values use the matching dataset audited 2026-09-12. These are sensitivity checks, not confidence intervals; ranking eligibility may also change.

Removed evidence	Recomputed score	Change
Artificial Analysis	55.7	+0.9
ARC Prize	57.0	+2.3
Datacurve	47.9	-6.8
LiveCodeBench	54.7	-0.0
Scale	55.8	+1.0
Terminal-Bench	55.8	+1.0
Cognition	55.2	+0.4
LiveBench	56.7	+2.0
SWE-rebench	54.1	-0.6
Epoch AI	55.1	+0.3
Vals AI	54.6	-0.1
Provider reports · Anthropic	54.7	-0.0
Provider reports · Moonshot AI	54.9	+0.2
Provider reports · Meta	54.7	-0.1
Provider reports · Mistral	54.7	+0.0
All provider reports	54.8	+0.1

Every benchmark family

Removed evidence	Recomputed score	Change
deepswe	47.9	-6.9
simpleqa	56.2	+1.4
arc-agi	55.8	+1.1
livebench-instructions	55.7	+0.9
harvey	55.7	+0.9
arc-agi-3	55.6	+0.9
mmmu-pro	55.5	+0.8
terminal-bench	55.5	+0.7
livebench-reasoning	54.1	-0.6
frontiermath	54.1	-0.6
swe-rebench	54.1	-0.6
vals-code-migration	54.1	-0.6
mmlu-pro	55.2	+0.5
frontiercode	55.2	+0.4
epoch-game-puzzles	54.3	-0.4
hle	55.1	+0.4
aime	54.4	-0.3
livebench-language	55.0	+0.3
livebench-data	54.4	-0.3
multi-swe-bench	54.4	-0.3
enterprise-ops	55.0	+0.3
swe-atlas-qna	54.5	-0.2
gpqa	54.5	-0.2
vals-excel-modeling	54.5	-0.2
vals-legal-research	54.6	-0.2
itbench	54.6	-0.2
automationbench	54.6	-0.1
omniscience	54.8	+0.1
swe-atlas-test-writing	54.6	-0.1
enigma-eval	54.8	+0.1
apex-agents	54.7	-0.1
aa-lcr	54.7	-0.1
livebench-coding	54.7	-0.0
tau-bench	54.7	-0.0
scicode	54.8	+0.0
ifbench	54.7	-0.0
swe-atlas-refactoring	54.7	-0.0
critpt	54.8	+0.0
gdp-pdf	54.7	-0.0
vals-finance-agent	54.8	+0.0
briefcase	54.7	-0.0
gmmlu	54.7	+0.0
livebench-math	54.7	+0.0
vibe-code	54.7	+0.0
math500	54.7	+0.0
gdpval	54.7	+0.0
livecodebench	54.7	+0.0
swe-bench-pro	54.7	-0.0
analyst-agent	54.7	+0.0

What evidence is missing from the fit?

Across all collected settings for this model: 246 observations, 176 contributing, 0 matched without graph weight and 70 excluded. Counts do not establish rank eligibility.

6 · Benchmark family or source protocol has not been reviewed
35 · Supporting evidence outside the reviewed capability core
28 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately
1 · Provider-sourced OSWorld task revision is unverified; cannot join a known-version comparison.

Inspect every observation and exclusion →

Capability profile

One configuration, seven capabilities

Comparison score · 0–100

AGPT-5.6 Terra · Max4 of 7 capabilities supported

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

Agentic
A59.3Supported
Hard reasoning
A62.5Supported
Coding
A62.9Supported
Human pref
Unknown
Knowledge
A33.3Supported
Multimodal
A46.5Preliminary
Long context
A70.2Preliminary

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.

Compare this configuration →Explore the model evidence profile →

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
AA-Briefcase EloArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Briefcase 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 2026-09-12.	1,014.49 Elo Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	1,040.72 Elo Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	1,201.36 Elo Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	1,337.28 Elo Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	1,330 Elo Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	946.69 Elo Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-non-reasoning
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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	38.94% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	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 2026-09-12.	71.33% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	74% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	77.67% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	79% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	83% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	58.67% Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-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 2026-09-12.	29.11% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	33.54% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	42.02% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	47.14% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	59.65% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	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 · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	9.43% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	17.43% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	22.86% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	27.14% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	30% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	2% Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-non-reasoning
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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	38.5% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	Not reported
GDP.pdf: All-passArtificial 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 2026-09-12.	17.6% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	17% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	20.8% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	24.6% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	24% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	Not reported
GDPval-AA v2 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 a human baseline of 1,000 · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-09-12.	1,177.99 Elo Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	1,318.34 Elo Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	1,415.36 Elo Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	1,478.6 Elo Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	1,476.96 Elo Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	1,168.63 Elo Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-non-reasoning
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 2026-09-12.	84.34% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	87.17% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	89.6% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	90.81% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	92.53% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	74.65% Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-non-reasoning
Harvey LAB-AA: Criterion Pass RateArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of atomic pass/fail rubric criteria the deliverables satisfy, graded by a single LLM judge · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	85.18% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	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 2026-09-12.	29.19% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	33.27% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	38.51% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	41.89% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	42.91% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	11.4% Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-non-reasoning
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 2026-09-12.	59.66% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	62.18% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	64.42% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	66.26% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	71.22% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	51.04% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	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 2026-09-12.	8.89% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	12.22% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	12.22% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	21.67% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	31.67% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	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 2026-09-12.	76.13% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	76.76% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	79.08% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	79.48% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	80.69% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	66.71% Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-non-reasoning
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 2026-09-12.	-6.83 Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	-5.13 Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	-3.47 Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	-2.98 Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	0.05 Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	-23.22 Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-non-reasoning
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 2026-09-12.	43.73% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	44.6% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	45.48% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	45.52% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	46.8% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	36.82% Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-non-reasoning
SciCodeArtificial 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 2026-09-12.	49.88% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	50.46% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	52.43% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	52.31% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	54.98% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	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 2026-09-12.	60.53% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	72.81% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	78.36% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	80.41% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	86.26% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	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 2026-09-12.	18.76% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	25.57% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	28.66% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	29.69% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	40.21% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	15.67% Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-non-reasoning
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 2026-09-12.	43.94% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	Not reported	57.58% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	62.88% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	57.58% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	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 2026-09-12.	62.55% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	72.28% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	75.66% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	80.15% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	88.01% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	Not reported	56.18% Reported settings GPT-5.6 Terra (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-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 2026-09-12.	1.52% Reported settings GPT-5.6 Terra (low); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-low	1.01% Reported settings GPT-5.6 Terra (medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-medium	1.52% Reported settings GPT-5.6 Terra (high); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-high	10.1% Reported settings GPT-5.6 Terra (xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra-xhigh	35.35% Reported settings GPT-5.6 Terra (max); Model source: https://artificialanalysis.ai/models/gpt-5-6-terra	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 2026-09-12.	18.75% Reported settings GPT-5.6 Terra (Low); Source model ID: openai-gpt-5-6-terra-low; 120 tasks	37.5% Reported settings GPT-5.6 Terra (Medium); Source model ID: openai-gpt-5-6-terra-medium; 120 tasks	67.08% Reported settings GPT-5.6 Terra (High); Source model ID: openai-gpt-5-6-terra-high; 120 tasks	74.17% Reported settings GPT-5.6 Terra (XHigh); Source model ID: openai-gpt-5-6-terra-xhigh; 120 tasks	83.9% Reported settings GPT-5.6 Terra (Max); Source model ID: openai-gpt-5-6-terra-max; 120 tasks	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 2026-09-12.	24.1% Reported settings gpt-5-6-terra; Reported effort: low; 113 unique tasks; 449 attempts; 4 runs	35.1% Reported settings gpt-5-6-terra; Reported effort: medium; 113 unique tasks; 450 attempts; 4 runs	53.8% Reported settings gpt-5-6-terra; Reported effort: high; 113 unique tasks; 452 attempts; 4 runs	60.2% Reported settings gpt-5-6-terra; Reported effort: xhigh; 113 unique tasks; 452 attempts; 4 runs	69.6% Reported settings gpt-5-6-terra; Reported effort: max; 113 unique tasks; 451 attempts; 4 runs	Not reported	Not reported
Terminal-Bench 2.1Terminal-Bench · Codex Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	78.43% Reported settings GPT-5.6 Terra; Reported effort: max; 445 trials; recorded 2026-06-26	Not reported	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 2026-09-12.	0.01% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-5-6-terra; variant low; public-demo task scores are excluded	0.08% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-5-6-terra; variant medium; public-demo task scores are excluded	0.49% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-5-6-terra; variant high; public-demo task scores are excluded	Not reported	0.8% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-5-6-terra; variant max; public-demo task scores are excluded	Not reported	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 2026-09-12.	37.79% 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	47.66% 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 0.0007	50.86% 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 0.0007	52.63% 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 0.0007	55.84% 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 0.0007	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 2026-09-12.	24.11% 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	33.88% 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 0.001	36.89% 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	38.75% 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 0.001	41.31% 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 0.001	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	21.5% Reported settings Terminal-Bench 4.0.0; 66 tasks; Codex; effort max; reported margin ±3.3 percentage points, confidence level unspecified; release date 2026-06-26; board checked 2026-09-07	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	98% Reported settings LiveBench official task evaluation; livebench-2026-06-25-AMPS_Hard; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	80.44% Reported settings LiveBench official task evaluation; livebench-2026-06-25-code_completion; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	76.06% Reported settings LiveBench official task evaluation; livebench-2026-06-25-code_generation; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	93% Reported settings LiveBench official task evaluation; livebench-2026-06-25-connections; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	87.67% Reported settings LiveBench official task evaluation; livebench-2026-06-25-consecutive_events; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · integrals with game · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	95% Reported settings LiveBench official task evaluation; livebench-2026-06-25-integrals_with_game; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · javascript · 2026-06-25LiveBench · LiveBench Mini-SWE-Agent, 250-step limit 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	68.18% Reported settings LiveBench Mini-SWE-Agent, 250-step limit; livebench-2026-06-25-javascript; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · logic with navigation · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	76% Reported settings LiveBench official task evaluation; livebench-2026-06-25-logic_with_navigation; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · math comp · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	95.1% Reported settings LiveBench official task evaluation; livebench-2026-06-25-math_comp; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · olympiad · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	91.54% Reported settings LiveBench official task evaluation; livebench-2026-06-25-olympiad; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · paraphrase · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	57.03% Reported settings LiveBench official task evaluation; livebench-2026-06-25-paraphrase; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · plot unscrambling · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	69.68% Reported settings LiveBench official task evaluation; livebench-2026-06-25-plot_unscrambling; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · python · 2026-06-25LiveBench · LiveBench Mini-SWE-Agent, 250-step limit 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	50% Reported settings LiveBench Mini-SWE-Agent, 250-step limit; livebench-2026-06-25-python; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · simplify · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	61.43% Reported settings LiveBench official task evaluation; livebench-2026-06-25-simplify; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · spatial · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	100% Reported settings LiveBench official task evaluation; livebench-2026-06-25-spatial; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · story 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	66.83% Reported settings LiveBench official task evaluation; livebench-2026-06-25-story_generation; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · summarize · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	73.17% Reported settings LiveBench official task evaluation; livebench-2026-06-25-summarize; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · tablejoin · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	50.25% Reported settings LiveBench official task evaluation; livebench-2026-06-25-tablejoin; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · tablereformat · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	100% Reported settings LiveBench official task evaluation; livebench-2026-06-25-tablereformat; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · theory of mind · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	86.54% Reported settings LiveBench official task evaluation; livebench-2026-06-25-theory_of_mind; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · typescript · 2026-06-25LiveBench · LiveBench Mini-SWE-Agent, 250-step limit 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	46.67% Reported settings LiveBench Mini-SWE-Agent, 250-step limit; livebench-2026-06-25-typescript; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · typos · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	86% Reported settings LiveBench official task evaluation; livebench-2026-06-25-typos; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
LiveBench · zebra puzzle · 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	100% Reported settings LiveBench official task evaluation; livebench-2026-06-25-zebra_puzzle; gpt-5.6-terra-max; checkpoint=gpt-5.6-terra	Not reported	Not reported
Epoch · Chess Puzzles · v1.1.6Epoch AI · Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	22% Reported settings Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match; model=gpt-5.6-terra_low; run=JzDofpJXt6d4f5kbRMSEmw; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported	Not reported	54% Reported settings Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match; model=gpt-5.6-terra_max; run=CQhcUcXfPPJKuyt57AHZw5; mean accuracy; https://epoch.ai/data/benchmarks.csv	5% Reported settings Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match; model=gpt-5.6-terra_none; run=LmRxqPrS2Gm6Gq2cSNc92G; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported
Epoch · FrontierMath Tier 4 · v2.0.0Epoch AI · Epoch AI Inspect; task 2.0.0; scorer verification_code Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. FrontierMath was developed with OpenAI funding and unequal access to part of the question set; this result is from Epoch's evaluation. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	70.73% Reported settings Epoch AI Inspect; task 2.0.0; scorer verification_code; model=gpt-5.6-terra_max; run=ZUygpkXqg3NjchXXfedAWV; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported
Epoch · FrontierMath Tiers 1–3 · v2.0.0Epoch AI · Epoch AI Inspect; task 2.0.0; scorer verification_code Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. FrontierMath was developed with OpenAI funding and unequal access to part of the question set; this result is from Epoch's evaluation. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	85.96% Reported settings Epoch AI Inspect; task 2.0.0; scorer verification_code; model=gpt-5.6-terra_max; run=J6evjTMgHkrun8txnEvzdN; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported
Epoch · GPQA diamond · v1.0.11Epoch AI · Epoch AI Inspect; task 1.0.11; scorer choice Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	87.37% Reported settings Epoch AI Inspect; task 1.0.11; scorer choice; model=gpt-5.6-terra_low; run=JcuoAWhcnngPEZBPmiUyLF; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported	Not reported	93.31% Reported settings Epoch AI Inspect; task 1.0.11; scorer choice; model=gpt-5.6-terra_max; run=4bLcFNu5zF6YuZf3qVKoHE; mean accuracy; https://epoch.ai/data/benchmarks.csv	77.27% Reported settings Epoch AI Inspect; task 1.0.11; scorer choice; model=gpt-5.6-terra_none; run=F7v3JtKFb27j2YVmsndLM5; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported
Epoch · Mystery Game Puzzles · v1.0.4 · Tool submissionEpoch AI · Epoch AI Inspect; task 1.0.4; scorer agent_submitted_next_move_scorer Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	19% Reported settings Epoch AI Inspect; task 1.0.4; scorer agent_submitted_next_move_scorer; model=gpt-5.6-terra_low; run=iE9NNTJWbGpRia2fUVGkPf; mean accuracy; https://epoch.ai/data/benchmarks.csv	10% Reported settings Epoch AI Inspect; task 1.0.4; scorer agent_submitted_next_move_scorer; model=gpt-5.6-terra_medium; run=WcCQQSHTFAhXk6qCWEeSLh; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported	35% Reported settings Epoch AI Inspect; task 1.0.4; scorer agent_submitted_next_move_scorer; model=gpt-5.6-terra_max; run=MFTd4BUbtzbfcCky6d79d4; mean accuracy; https://epoch.ai/data/benchmarks.csv	14% Reported settings Epoch AI Inspect; task 1.0.4; scorer agent_submitted_next_move_scorer; model=gpt-5.6-terra_none; run=8grv57UmR89LXkEYhyssr9; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported
Epoch · OTIS Mock AIME 2024-2025 · v1.1.6Epoch AI · Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	88.89% Reported settings Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match; model=gpt-5.6-terra_low; run=Z6S3nMC2tnkpKUixCPLiDj; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported	Not reported	99.72% Reported settings Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match; model=gpt-5.6-terra_max; run=jUmmCNezoW7jWiK72SVCaP; mean accuracy; https://epoch.ai/data/benchmarks.csv	53.33% Reported settings Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match; model=gpt-5.6-terra_none; run=cjUVqF2qQJ4qhFKBU5BQpW; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported
Epoch · SimpleQA Verified · v1.0.0Epoch AI · Epoch AI Inspect; task 1.0.0; scorer simpleqa_scorer Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	43.2% Reported settings Epoch AI Inspect; task 1.0.0; scorer simpleqa_scorer; model=gpt-5.6-terra_max; run=DienatPbgfgn3EzM7Gvzxr; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported
Vals AI · Finance Agent v2Vals AI · Vals FAB v2 shared six-tool harness; two-hour task limit Benchmark details Overall weighted rubric accuracy, averaged over three runs on 450 tasks. Dealbreaker failures and timeouts score zero. Related categories and all-pass scores are not additional votes. Three-model judge panel; private test set. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	54.44% Reported settings {"checkpoint":"gpt-5.6-terra","evaluatorModel":"openai/gpt-5.6-terra","sourceUpdatedOn":"2026-09-10","reasoning_effort":"max","compute_effort":null,"reasoning":null,"temperature":null,"top_p":null,"max_output_tokens":128000,"verbosity":null,"provider":"OpenAI","harness":null}	Not reported	Not reported
Vals AI · Code Migration v1Vals AI · Vals Code Migration shared agent harness Benchmark details Hidden behavior-test pass rate, equally weighted across 40 repositories (30 CLI and 10 COBOL-to-Java). Anti-cheat failures score zero. Code-quality ratings and language splits are not additional votes. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	47.8% Reported settings {"checkpoint":"gpt-5.6-terra","evaluatorModel":"openai/gpt-5.6-terra","sourceUpdatedOn":"2026-09-10","reasoning_effort":"max","compute_effort":null,"reasoning":null,"temperature":null,"top_p":null,"max_output_tokens":128000,"verbosity":null,"provider":"OpenAI","harness":null}	Not reported	Not reported
Vals AI · Legal Research Bench v1Vals AI · Vals Legal Research shared five-tool harness; three-hour task limit Benchmark details All-pass accuracy on 208 private legal research tasks. Every expert rubric check must pass; GPT-5.4 judges responses. Partial-credit scores and practice-area splits are not additional votes. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	41.35% Reported settings {"checkpoint":"gpt-5.6-terra","evaluatorModel":"openai/gpt-5.6-terra","sourceUpdatedOn":"2026-09-10","reasoning_effort":"max","compute_effort":null,"reasoning":null,"temperature":null,"top_p":null,"max_output_tokens":128000,"verbosity":null,"provider":"OpenAI","harness":null}	Not reported	Not reported
Vals AI · Excel Modeling Benchmark v1Vals AI · Vals EMB shared workbook agent harness; 3.5-hour task limit Benchmark details Mean of seven equally weighted task categories. Template submissions use recalculated Excel cell accuracy; scratch submissions use rubric checks judged by GPT-5.4 Mini XHigh. Partial credit does not mean task completion. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	66.21% Reported settings {"checkpoint":"gpt-5.6-terra","evaluatorModel":"openai/gpt-5.6-terra","sourceUpdatedOn":"2026-09-10","reasoning_effort":"max","compute_effort":null,"reasoning":null,"temperature":null,"top_p":null,"max_output_tokens":128000,"verbosity":null,"provider":"OpenAI","harness":null}	Not reported	Not reported
Vals AI · MMLU-Pro v1Vals AI · Vals MMLU-Pro five-shot chain-of-thought; regex answer grading Benchmark details Mean accuracy across 14 subjects under Vals five-shot prompting. Subject splits share one family and are not additional votes. Vals labels this benchmark archived; it is historical corroboration, not ongoing frontier evaluation. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	86.66% Reported settings {"checkpoint":"gpt-5.6-terra","evaluatorModel":"openai/gpt-5.6-terra","sourceUpdatedOn":"2026-09-01","reasoning_effort":"xhigh","compute_effort":null,"reasoning":null,"temperature":null,"top_p":null,"max_output_tokens":128000,"verbosity":null,"provider":"OpenAI","harness":null}	Not reported	Not reported	Not reported
Vals AI · Terminal-Bench 2.1Vals AI · Vals Terminus 2 harness; pass@1 on 89 tasks Benchmark details Overall task pass rate under the shared Terminus 2 configuration. All task tests must pass. Difficulty splits share one family. Fable 5 and Opus 5 mixed-checkpoint fallback runs are excluded. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	77.53% Reported settings {"checkpoint":"gpt-5.6-terra","evaluatorModel":"openai/gpt-5.6-terra","sourceUpdatedOn":"2026-09-10","reasoning_effort":"max","compute_effort":null,"reasoning":null,"temperature":null,"top_p":null,"max_output_tokens":128000,"verbosity":null,"provider":"OpenAI","harness":null}	Not reported	Not reported
Vals AI · Vibe Code Bench v1.1Vals AI · Vals modified OpenHands harness; five hours or 1,000 turns per app Benchmark details Mean app test pass rate on the canonical v1.1 test set. Browser Use executes workflow tests; each test requires at least 90% of its substeps. Automated judging can introduce error. Alternate agents are excluded. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	74.59% Reported settings {"checkpoint":"gpt-5.6-terra","evaluatorModel":"openai/gpt-5.6-terra","sourceUpdatedOn":"2026-09-10","reasoning_effort":"max","compute_effort":null,"reasoning":null,"temperature":null,"top_p":null,"max_output_tokens":128000,"verbosity":null,"provider":"OpenAI","harness":"OpenHands"}	Not reported	Not reported
Vals AI · Harvey Legal Agent Benchmark v1Vals AI · Vals Harvey held-out legal file-work protocol; internet disabled Benchmark details Mean task resolution across two judges (GPT-5.5 Medium and Sonnet 4.6), requiring every criterion to pass. Criteria pass rates and practice-area splits do not add votes. Shared Harvey family; Fable 5 fallback run excluded. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	0.83% Reported settings {"checkpoint":"gpt-5.6-terra","evaluatorModel":"openai/gpt-5.6-terra","sourceUpdatedOn":"2026-09-10","reasoning_effort":"max","compute_effort":null,"reasoning":null,"temperature":null,"top_p":null,"max_output_tokens":128000,"verbosity":null,"provider":"OpenAI","harness":null}	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 · 7,424 results across 240 models · Collected 2026-09-12

GPT-5.6 Terra effort benchmarks

GPT-5.6 Terra

Capability score evidence

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context