Rankings / OpenAI

GPT 5.6 Sol

proprietary

released 2026-07-09

BenchAtlas Index

as of 2026-07-14

max effort

rank #1

14 families · 5 categories · high

xhigh effort

rank #2

13 families · 5 categories · high

high effort

rank #4

8 families · 5 categories · high

Benchmark evidence

142 results

agentic coding

Release 2026-06-25 tasks_counted=3 · livebench_version=2026-06-25	max effort	65.6 %	independent	LiveBench
Release 2026-06-25 tasks_counted=3 · livebench_version=2026-06-25	xhigh effort	56.6 %	independent	LiveBench
τ²-Bench	max effort	85.1 %	independent	Artificial Analysis
τ²-Bench	xhigh effort	84.8 %	independent	Artificial Analysis
τ²-Bench	high effort	83.3 %	independent	Artificial Analysis
τ²-Bench	medium effort	81.0 %	independent	Artificial Analysis
τ²-Bench	low effort	76.0 %	independent	Artificial Analysis
τ²-Bench subset=banking	max effort	33.0 %	independent	Artificial Analysis
τ²-Bench subset=banking	xhigh effort	32.6 %	independent	Artificial Analysis
τ²-Bench subset=banking	high effort	30.6 %	independent	Artificial Analysis
τ²-Bench subset=banking	medium effort	26.5 %	independent	Artificial Analysis
τ²-Bench subset=banking	low effort	24.4 %	independent	Artificial Analysis
τ²-Bench subset=banking	no reasoning	16.1 %	independent	Artificial Analysis

coding

Release 2026-06-25 tasks_counted=2 · livebench_version=2026-06-25	max effort	83.9 %	independent	LiveBench
Release 2026-06-25 tasks_counted=2 · livebench_version=2026-06-25	xhigh effort	81.8 %	independent	LiveBench
SciCode	high effort	56.9 %	independent	Artificial Analysis
SciCode	medium effort	56.5 %	independent	Artificial Analysis
SciCode	max effort	56.1 %	independent	Artificial Analysis
SciCode	xhigh effort	56.0 %	independent	Artificial Analysis
SciCode	low effort	55.4 %	independent	Artificial Analysis
SciCode	no reasoning	47.1 %	independent	Artificial Analysis

data analysis

Release 2026-06-25 tasks_counted=3 · livebench_version=2026-06-25	xhigh effort	80.3 %	independent	LiveBench
Release 2026-06-25 tasks_counted=3 · livebench_version=2026-06-25	max effort	79.8 %	independent	LiveBench

external indices

AA Coding Index	xhigh effort	78.3 points	independent	Artificial Analysis
AA Coding Index	max effort	77.4 points	independent	Artificial Analysis
AA Coding Index	high effort	77.2 points	independent	Artificial Analysis
AA Coding Index	medium effort	76.3 points	independent	Artificial Analysis
AA Coding Index	low effort	69.7 points	independent	Artificial Analysis
AA Coding Index	no reasoning	65.1 points	independent	Artificial Analysis
Intelligence Index v4.1	max effort	58.9 points	independent	Artificial Analysis
Intelligence Index v4.1	xhigh effort	57.7 points	independent	Artificial Analysis
Intelligence Index v4.1	high effort	55.9 points	independent	Artificial Analysis
Intelligence Index v4.1	medium effort	53.6 points	independent	Artificial Analysis
Intelligence Index v4.1	low effort	49.4 points	independent	Artificial Analysis
Intelligence Index v4.1	no reasoning	41.2 points	independent	Artificial Analysis
ECI	max effort	161.8 points	independent	Epoch AI Benchmarking Hub
ECI		161.8 points	independent	Epoch AI Benchmarking Hub
ECI	medium effort	161.8 points	independent	Epoch AI Benchmarking Hub
ECI	high effort	161.8 points	independent	Epoch AI Benchmarking Hub
ECI	low effort	161.8 points	independent	Epoch AI Benchmarking Hub
ECI	max effort	161.8 points	independent	Epoch AI Benchmarking Hub
ECI		161.8 points	independent	Epoch AI Benchmarking Hub
ECI	xhigh effort	161.8 points	independent	Epoch AI Benchmarking Hub
Vals Index	max effort	72.6 points	independent	Vals AI

factuality

SimpleQA Verified

max effort

71.6 %

independent

Epoch AI Benchmarking Hub

2026-07-09

human preference

Coding (style control)older version arena=text · category=coding · style_control=true	xhigh effort	1526.1 1504.2–1548.0	community	LMArena Leaderboard Dataset
Hard Prompts English (style control)older version arena=text · category=hard_prompts_english · style_control=true	xhigh effort	1522.5 1500.1–1544.8	community	LMArena Leaderboard Dataset
Industry Software And It Services (style control)older version arena=text · category=industry_software_and_it_services · style_control=true	xhigh effort	1508.7 1490.5–1527.0	community	LMArena Leaderboard Dataset
Hard Prompts (style control)older version arena=text · category=hard_prompts · style_control=true	xhigh effort	1504.8 1490.6–1519.0	community	LMArena Leaderboard Dataset
English (style control)older version arena=text · category=english · style_control=true	xhigh effort	1503.1 1485.3–1520.8	community	LMArena Leaderboard Dataset
Exclude Ties (style control)older version arena=text · category=exclude_ties · style_control=true	xhigh effort	1498.0 1482.8–1513.2	community	LMArena Leaderboard Dataset
Industry Business And Management And Financial Operations (style control)older version arena=text · category=industry_business_and_management_and_financial_operations · style_control=true	xhigh effort	1495.8 1469.7–1521.9	community	LMArena Leaderboard Dataset
Longer Query (style control)older version arena=text · category=longer_query · style_control=true	xhigh effort	1487.0 1468.7–1505.4	community	LMArena Leaderboard Dataset
Overall (style control) arena=text · category=overall · style_control=true	xhigh effort	1484.6 1473.2–1496.1	community	LMArena Leaderboard Dataset
Instruction Following (style control)older version arena=text · category=instruction_following · style_control=true	xhigh effort	1482.3 1462.2–1502.5	community	LMArena Leaderboard Dataset
Industry Writing And Literature And Language (style control)older version arena=text · category=industry_writing_and_literature_and_language · style_control=true	xhigh effort	1480.4 1457.9–1502.9	community	LMArena Leaderboard Dataset
Industry Entertainment And Sports And Media (style control)older version arena=text · category=industry_entertainment_and_sports_and_media · style_control=true	xhigh effort	1475.4 1451.2–1499.6	community	LMArena Leaderboard Dataset
Non English (style control)older version arena=text · category=non_english · style_control=true	xhigh effort	1463.9 1449.3–1478.6	community	LMArena Leaderboard Dataset

knowledge science

GPQA Diamond implementation=vals-ai	max effort	95.2 %	independent	Vals AI
GPQA Diamond implementation=artificial-analysis	max effort	94.1 %	independent	Artificial Analysis
GPQA Diamond	max effort	93.5 %	independent	Epoch AI Benchmarking Hub 2026-07-09
GPQA Diamond implementation=artificial-analysis	xhigh effort	93.1 %	independent	Artificial Analysis
GPQA Diamond implementation=artificial-analysis	high effort	92.8 %	independent	Artificial Analysis
GPQA Diamond implementation=artificial-analysis	medium effort	92.6 %	independent	Artificial Analysis
GPQA Diamond implementation=artificial-analysis	low effort	89.8 %	independent	Artificial Analysis
GPQA Diamond implementation=artificial-analysis	no reasoning	79.0 %	independent	Artificial Analysis
Humanity's Last Exam implementation=artificial-analysis	max effort	47.2 %	independent	Artificial Analysis
Humanity's Last Exam implementation=artificial-analysis	xhigh effort	44.7 %	independent	Artificial Analysis
Humanity's Last Exam implementation=artificial-analysis	high effort	44.1 %	independent	Artificial Analysis
Humanity's Last Exam implementation=artificial-analysis	medium effort	39.7 %	independent	Artificial Analysis
Humanity's Last Exam implementation=artificial-analysis	low effort	36.6 %	independent	Artificial Analysis
Humanity's Last Exam implementation=artificial-analysis	no reasoning	15.8 %	independent	Artificial Analysis
MMLU-Pro implementation=vals-ai	max effort	89.1 %	independent	Vals AI

language

Release 2026-06-25 tasks_counted=3 · livebench_version=2026-06-25	max effort	87.7 %	independent	LiveBench
Release 2026-06-25 tasks_counted=3 · livebench_version=2026-06-25	xhigh effort	85.9 %	independent	LiveBench

long context instruction

AA-LCR	max effort	73.7 %	independent	Artificial Analysis
AA-LCR	xhigh effort	71.0 %	independent	Artificial Analysis
AA-LCR	medium effort	68.7 %	independent	Artificial Analysis
AA-LCR	high effort	68.3 %	independent	Artificial Analysis
AA-LCR	low effort	67.7 %	independent	Artificial Analysis
AA-LCR	no reasoning	55.3 %	independent	Artificial Analysis
IFBench	max effort	72.7 %	independent	Artificial Analysis
IFBench	xhigh effort	71.0 %	independent	Artificial Analysis
IFBench	medium effort	69.6 %	independent	Artificial Analysis
IFBench	high effort	69.2 %	independent	Artificial Analysis
IFBench	low effort	66.5 %	independent	Artificial Analysis
Release 2026-06-25 tasks_counted=4 · livebench_version=2026-06-25	max effort	71.8 %	independent	LiveBench
Release 2026-06-25 tasks_counted=4 · livebench_version=2026-06-25	xhigh effort	67.4 %	independent	LiveBench

multimodal

MMMU

implementation=vals-ai

max effort

88.8 %

independent

Vals AI

professional

CorpFin	max effort	64.4 %	independent	Vals AI
LegalBench	max effort	87.0 %	independent	Vals AI
TaxEval	max effort	74.8 %	independent	Vals AI

reasoning math

ARC-AGI-1older version split=public_eval · model_type=CoT	max effort	99.0 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	xhigh effort	98.8 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	xhigh effort	97.5 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	high effort	97.3 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	high effort	97.0 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	max effort	96.5 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	medium effort	93.9 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	max effort	93.8 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	medium effort	92.5 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	max effort	92.5 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	xhigh effort	91.9 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	xhigh effort	90.0 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	high effort	86.1 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	high effort	85.4 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	low effort	84.5 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	low effort	74.5 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	medium effort	70.7 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	medium effort	67.1 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	low effort	42.5 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	low effort	38.5 %	independent	ARC Prize Leaderboard
ARC-AGI-3older version split=semi_private · model_type=CoT	max effort	7.8 %	independent	ARC Prize Leaderboard
ARC-AGI-3older version split=semi_private · model_type=CoT	xhigh effort	7.0 %	independent	ARC Prize Leaderboard
ARC-AGI-3older version split=semi_private · model_type=CoT	high effort	2.1 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	max effort	1.5 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	max effort	1.4 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-3older version split=semi_private · model_type=CoT	medium effort	1.1 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	xhigh effort	1.0 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	xhigh effort	1.0 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	high effort	0.8 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	high effort	0.7 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	max effort	0.5 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	medium effort	0.5 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	medium effort	0.5 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	xhigh effort	0.4 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	max effort	0.4 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-3older version split=semi_private · model_type=CoT	low effort	0.3 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	low effort	0.3 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	low effort	0.3 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	high effort	0.3 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	xhigh effort	0.3 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	high effort	0.2 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	medium effort	0.2 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	low effort	0.2 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	medium effort	0.2 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	low effort	0.1 usd_per_task	independent	ARC Prize Leaderboard
Release 2026-06-25 tasks_counted=4 · livebench_version=2026-06-25	max effort	96.2 %	independent	LiveBench
Release 2026-06-25 tasks_counted=4 · livebench_version=2026-06-25	xhigh effort	95.5 %	independent	LiveBench
Release 2026-06-25 tasks_counted=4 · livebench_version=2026-06-25	max effort	91.7 %	independent	LiveBench
Release 2026-06-25 tasks_counted=4 · livebench_version=2026-06-25	xhigh effort	90.2 %	independent	LiveBench
OTIS Mock AIME 2024–2025	max effort	100.0 %	independent	Epoch AI Benchmarking Hub 2026-07-09

Agent + model results

systems, not bare-model scores

agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench 2.1	89.5 %	independent	Artificial Analysis
agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench 2.1	88.0 %	independent	Artificial Analysis
agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench 2.1	87.3 %	independent	Artificial Analysis
agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench 2.1	86.1 %	independent	Artificial Analysis
agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench 2.1	76.8 %	independent	Artificial Analysis
agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench 2.1	74.2 %	independent	Artificial Analysis
agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench Hard	65.9 %	independent	Artificial Analysis
agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench Hard	62.9 %	independent	Artificial Analysis
agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench Hard	62.1 %	independent	Artificial Analysis
agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench Hard	61.4 %	independent	Artificial Analysis
agent + model Artificial Analysis harness + GPT 5.6 Sol	Terminal-Bench Hard	60.6 %	independent	Artificial Analysis

These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.