Benchmark Almanac

GPT-5.6 Terra

GPT-5.6 Terra places in the top three on 9 benchmarks.

OpenAI Released July 9th, 2026 Compare with others
247
Benchmarks with a score
0
Ranked first, out of five or more models
9
In the top three
OpenAI Jul 2026

GPT-5.6 Terra

Closed weights

160.7 #17 of 256
Epoch Capabilities Index
Self-Reported Index 158.7 #40 of 404
  • MMMU-Pro 82% #8 /90
  • DeepSWE 69.6% #6 /28
  • AutomationBench 23.6% #13 /42

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

BenchmarkScoreRankSource
HLE Extended 70.7% #2 of 31 Official leaderboard
AA-LCR Extended 74.4% #4 of 49 Official leaderboard
LMArena (Elo) 1,468 #41 of 335 Official leaderboard
Creative Writing v3 79.5 #16 of 120 Official leaderboard
Opening Lines Arena 1,602 #2 of 15 Official leaderboard
StudyArena Text 1,276 #12 of 85 Official leaderboard
Wer wird Millionär? LLM Benchmark 183333.0 #13 of 77 Official leaderboard
SEA-HELM 75.5 #12 of 71 Official leaderboard
OpenAI MRCR v2 67.9% #8 of 47 Official leaderboard
Longform Creative Writing (EQ-Bench) 78.0 #22 of 119 Official leaderboard
ELI5 Arena 1,591 #5 of 18 Official leaderboard
Document Arena 1,474 #11 of 39 Official leaderboard
Global MMLU Lite 92.7% #15 of 43 Official leaderboard
BullshitBench V2 54% #46 of 129 Official leaderboard
Last Translation Benchmark 12.2 #13 of 36 Official leaderboard
FACTS Parametric 56.7% #17 of 45 Official leaderboard
SimpleQA Verified 41.5% #22 of 56 Official leaderboard
Joke Arena 1,516 #10 of 25 Official leaderboard
BullshitBench 34.5% #44 of 107 Official leaderboard
FACTS Search V2 81.7% #9 of 21 Official leaderboard
Extended NYT Connections 78.4% #46 of 103 Official leaderboard
Continual Learning Exam 33.2% #5 of 11 Official leaderboard
DesignArena ASCII Art 1,171 #36 of 77 Official leaderboard
EQ-Bench 4 1,220 #15 of 32 Official leaderboard
HUMAINE 29 #24 of 50 Official leaderboard
FACTS Parametric V2 57.6% #13 of 27 Official leaderboard
GraphWalks 76.9% #13 of 27 Lab model card
FACTS Benchmark Suite 61.3% #15 of 23 Official leaderboard
FACTS Multimodal 39.9% #17 of 23 Official leaderboard
FACTS Grounding v2 60.5% #19 of 25 Official leaderboard
Corp Speak Arena 1,451 #17 of 22 Official leaderboard

Math and science

BenchmarkScoreRankSource
ProofBench 71% #4 of 59 Official leaderboard
OTIS Mock AIME 2024-2025 99.7% #16 of 212 Epoch AI run
GPQA Diamond 93.3% #21 of 229 Epoch AI run
ARC-AGI 96.5% #14 of 95 Official leaderboard
FrontierMath Tiers 1-3 v2 86% #13 of 86 Official leaderboard
ARC-AGI-2 83.9% #17 of 98 Official leaderboard
Verifying the Verifiers (Project APE CRED benchmark) 97% #4 of 23 Official leaderboard
DTBench 93.3% #33 of 180 Official leaderboard
FrontierMath Tier 4 v2 70.7% #16 of 68 Official leaderboard
Surface Evolver Bench 83.8% #6 of 25 Official leaderboard
ProofBench v1.1 74% #19 of 52 Official leaderboard
FrontierChallenge 15.5% #5 of 11 Official leaderboard
SimpleBench 48.9% #61 of 106 Official leaderboard
Terminal-Bench-Science 8.57% #11 of 17 Official leaderboard
FrontierPhysics 3.60% #15 of 22 Official leaderboard
Human Behavior Bench 77.8 #8 of 11 Official leaderboard

Coding

BenchmarkScoreRankSource
WeirdML 78.3% #11 of 112 Official leaderboard
SWE-bench Verified Extended 78.7% #5 of 44 Official leaderboard
ALE-Bench 2,154 #15 of 126 Official leaderboard
IOI (Vals) 65.3% #7 of 55 Official leaderboard
IOI v2 (Vals) 87.6% #7 of 45 Official leaderboard
Code Migration 47.8% #13 of 78 Official leaderboard
DeepSWE 69.6% #6 of 28 Official leaderboard
ReactBench 39% #6 of 28 Official leaderboard
AndroidBench 86.8% #7 of 32 Official leaderboard
SWE-Bench Pro 63.4% #19 of 81 Lab model card
DesignArena Website 1,261 #46 of 151 Official leaderboard
DesignArena SVG 1,209 #34 of 103 Official leaderboard
WebDev Arena 1,521 #36 of 105 arena.ai, via Epoch AI
Vibe Code Bench 74.6% #45 of 120 Official leaderboard
DesignArena UI Component 1,250 #53 of 139 Official leaderboard
DesignArena Data Visualization 1,237 #46 of 118 Official leaderboard
DesignArena Game Dev 1,220 #56 of 143 Official leaderboard
Code Arena 1,520 #53 of 128 Official leaderboard
DesignArena Code 1,230 #61 of 143 Official leaderboard
SWE-Bench Pro V2 Hard 86.3% #8 of 18 Official leaderboard
SWE Atlas Codebase QnA Extended 16% #18 of 38 Official leaderboard
Senior SWE-Bench 27.4% #11 of 23 Official leaderboard
Image-to-WebDev Arena 1,539 #29 of 55 Official leaderboard
MirrorCode 41.1% #21 of 39 Epoch AI run
SWE-sweep 1.50% #6 of 10 Official leaderboard
τ^τ-bench 18% #3 of 5 Official leaderboard
DesignArena 3D Design 1,112 #87 of 136 Official leaderboard
DesignArena Image to HTML 1,184 #34 of 51 Official leaderboard
Artificial Analysis Coding Agent Index v1.1 77.4 #2 of 3 Lab model card
DesignArena Agentic Overall 1,083 #53 of 74 Official leaderboard
Vibe Code Bench 1-100 14.8% #19 of 26 Official leaderboard
CursorBench 41.3% #16 of 21 cursor.com, via Epoch AI
DesignArena Full Stack 1,041 #39 of 49 Official leaderboard
APEX-SWE 31.1% #52 of 65 Official leaderboard
DesignArena React Native 1,099 #45 of 56 Official leaderboard
CursorBench 4.0 41.3% #17 of 21 Official leaderboard
DesignArena Frontend Web App 1,101 #43 of 53 Official leaderboard
DesignArena Android (Kotlin) 1,056 #39 of 46 Official leaderboard
SWE-Bench Pro V2 92.4% #16 of 18 Official leaderboard
Vision2Web 67.5% #9 of 10 Official leaderboard
DesignArena Agentic Game Dev 1,032 #26 of 27 Official leaderboard
DesignArena Image to Frontend 1,081 #30 of 31 Official leaderboard

Agents and tools

BenchmarkScoreRankSource
LLM Chess 1,293 #13 of 123 Official leaderboard
BALROG 53.2% #6 of 33 Official leaderboard
BrowseComp 87.5% #15 of 81 Lab model card
RuneBench 5.88 #15 of 76 Official leaderboard
SnakeBench 30 #42 of 188 Official leaderboard
Game Arena Poker 7.27 #7 of 31 Official leaderboard
LisanBench 3302.7 #34 of 133 Official leaderboard
Chess Puzzles 47% #25 of 93 Epoch AI run
GAIA 71.8% #18 of 64 Official leaderboard
Mystery Game Puzzles 35% #26 of 84 Epoch AI run
AutomationBench 23.6% #13 of 42 Official leaderboard
Toolathlon 53.1% #9 of 29 Lab model card
GameHorizon 70.7 #14 of 37 Official leaderboard
PinchBench 2.0 75.9% #22 of 57 Official leaderboard
Terminal-Bench 2.1 Extended 36.4% #17 of 44 Official leaderboard
OSWorld-Science 31.8% #5 of 11 Official leaderboard
Terminal-Bench 2.1 78.4% #6 of 13 Official leaderboard
BabaIsBench 88% #16 of 32 Official leaderboard
CAR-bench 56.7% #6 of 12 Official leaderboard
BrowseComp Extended 25.3% #22 of 42 Official leaderboard
Game Arena Chess 725 #25 of 42 Official leaderboard
Terminal-Bench 4.0 21.5% #24 of 40 Official leaderboard
Terminal-Bench 4.1 21.5% #26 of 43 Official leaderboard
ApprenticeBench 16% #11 of 18 Official leaderboard
Agent Arena -1.85% #35 of 56 Official leaderboard
ARC-AGI-3 0.80% #17 of 27 Official leaderboard
Terminal-Bench 3.0 20.8% #7 of 11 Official leaderboard
CLASHarness 20% #9 of 14 Official leaderboard
Game Arena Word Association 177 #26 of 40 Official leaderboard
Game Arena Word Art 295 #26 of 39 Official leaderboard
Game Arena Checkers 746 #28 of 41 Official leaderboard
Game Arena Lines of Action 310 #27 of 39 Official leaderboard
Bench on the Clocktower 48.7% #7 of 10 Official leaderboard
TextQuests 20.2 #20 of 28 Official leaderboard
Game Arena Clobber 160 #28 of 39 Official leaderboard
OSWorld 2.0 50.2% #14 of 19 Lab model card
Game Arena Go 512 #29 of 39 Official leaderboard
Kaggle Game Arena 198 #32 of 42 Official leaderboard
Game Arena Dark Hex 129 #32 of 41 Official leaderboard
Game Arena Dots and Boxes 603 #32 of 40 Official leaderboard
DuelLab GameBench 2 53.5 #47 of 57 Official leaderboard
Brood War Bench 50% #5 of 6 Official leaderboard
Game Arena Bargaining 132 #34 of 39 Official leaderboard
Game Arena Coin Game 206 #34 of 39 Official leaderboard
Game Arena Reversi 148 #34 of 39 Official leaderboard
Game Arena Ultimate Tic-Tac-Toe 57 #34 of 39 Official leaderboard
DuetBench 952 #8 of 9 Official leaderboard
Game Arena Nine Men's Morris 262 #35 of 39 Official leaderboard
Game Arena Four in a Row 92 #41 of 43 Official leaderboard

Professional work

BenchmarkScoreRankSource
TaxEval v2 76.2% #7 of 129 Official leaderboard
ProfBench 57.2% #11 of 72 Official leaderboard
APEX-1 69.5% #11 of 71 Official leaderboard
LegalBench 85.1% #25 of 138 Official leaderboard
Vending-Bench Arena $9,191 #6 of 32 Official leaderboard
TRLawBench 9.50 #7 of 34 Official leaderboard
Vals Index v2 59.6% #15 of 65 Official leaderboard
CorpFin v2 65.3% #32 of 119 Official leaderboard
BenchCAD 78.2% #3 of 11 Lab model card
MortgageTax 67.3% #24 of 87 Official leaderboard
Management Consulting Tasks (OpenAI) 37.2% #2 of 7 Official leaderboard
Vending-Bench 2 $7,343 #20 of 67 Official leaderboard
Excel Modeling Benchmark 66.2% #23 of 76 Official leaderboard
Agents' Last Exam 28% #10 of 31 Official leaderboard
Quant-Arena 47.6 #3 of 9 Official leaderboard
Vals Finance Agent v2 54.4% #27 of 80 Official leaderboard
APEX-Agents 1.1 71.8% #22 of 65 Official leaderboard
PolicyBench 87.6% #17 of 50 Official leaderboard
Legal Research Bench 41.4% #32 of 93 Official leaderboard
APEX-Agents 58.2% #21 of 61 Collected by Epoch AI, source not named
Tax Agent Bench 30% #35 of 84 Official leaderboard
GDPval Extended 1.30% #18 of 43 Official leaderboard
APEX-Accounting 59.8% #21 of 49 Official leaderboard
GDPval-AA v2 1,593 #12 of 27 Lab model card
MedCode 43.4% #43 of 96 Official leaderboard
Vals Index v2.1 53.1% #22 of 49 Official leaderboard
MedScribe 82.9% #47 of 97 Official leaderboard
BigFinanceBench 51% #1 of 2 Lab model card
DAYJOB: Healthcare 3.60% #20 of 37 Official leaderboard
Public Benefits Bench 62.4% #28 of 50 Official leaderboard
HealthBench Hard 32.7% #12 of 21 Lab model card
LifeSci Diamond 1.80% #11 of 18 Official leaderboard
HealthBench Consensus 95.1% #11 of 17 Lab model card
YC-Bench $636,144 #43 of 59 Official leaderboard
MedXpertQA MM Extended 51% #30 of 41 Official leaderboard
HealthBench 57% #23 of 31 Lab model card
Harvey LAB (Legal Agent Benchmark) 0.83% #60 of 80 Official leaderboard
RCM-Bench-1 87% #15 of 20 Official leaderboard
DAYJOB: Finance 3.30% #30 of 39 Official leaderboard
HealthBench Professional 57.7% #20 of 26 Lab model card
CHI-Bench 13.3% #18 of 23 Official leaderboard
DesignArena Python-PPTX Slides 1,088 #29 of 36 Official leaderboard
DesignArena Spreadsheet 1,062 #30 of 35 Official leaderboard
DesignArena HTML Slides 1,002 #26 of 28 Official leaderboard
DesignArena PDF 1,032 #36 of 37 Official leaderboard
Parametric CAD Bench 17.2% #11 of 11 Official leaderboard
Genspark Slides Benchmark 48.8% #10 of 10 Official leaderboard

Vision, audio and robotics

BenchmarkScoreRankSource
MMMU-Pro 82% #8 of 90 Lab model card
MMMU-Pro Extended 47.4% #7 of 38 Official leaderboard
Vision Arena 1,266 #30 of 136 Official leaderboard
VoxelBench 1,921 #14 of 56 Official leaderboard
Chartography 34% #11 of 43 Official leaderboard
ExtractBench 92.3% #7 of 27 Official leaderboard
ZeroBench 19% #23 of 76 Official leaderboard
vCX Benchmark 90.1% #13 of 42 Official leaderboard
BVB 68.8 #16 of 45 Official leaderboard
Furniture Assembly Benchmark 54.2% #13 of 35 Official leaderboard
ParseBench 67.1% #18 of 44 Official leaderboard
CharXiv Extended 81.1% #17 of 38 Official leaderboard
ArtifactArena 1,093 #10 of 21 Official leaderboard
RoTA-Bench 8.00% #4 of 8 Official leaderboard
Blueprint-Bench 2 30.8% #22 of 38 Official leaderboard
4DCodeBench 0.55 #10 of 16 Official leaderboard
GDP.pdf 24.7% #12 of 17 Lab model card
1H-VideoQA 67.3% #15 of 20 Official leaderboard
Humanity's Sixth Sense 22.7% #32 of 34 Official leaderboard
DesignArena Code to Video 972 #28 of 29 Official leaderboard
SplatBench 1.20 #8 of 8 Official leaderboard

Safety and security

BenchmarkScoreRankSource
Human Pathogen Capabilities Test v2 (HPCT v2) 70% #6 of 71 Official leaderboard
Human Pathogen Capabilities Test 68.2% #8 of 86 Official leaderboard
Tacit Knowledge and Troubleshooting 84.1% #3 of 30 Lab model card
World-Class Bio 59.8% #9 of 74 Official leaderboard
Molecular Biology Capabilities Test 56.9% #15 of 90 Official leaderboard
GeneBench Pro 28.5% #2 of 12 Official leaderboard
nocot-bench (NCRI) 89.0 #37 of 185 Official leaderboard
DistressBench 76.4% #10 of 42 Official leaderboard
Agent JSK 100% #2 of 7 Lab model card
Prompt injection attacks in search and function calling 94.6% #1 of 3 Lab model card
CyberBench v1.1 (Vals) 72.1% #16 of 44 Official leaderboard
SRE-Bench 9.54% #13 of 35 Official leaderboard
BioTIER Refuse 71.7% #32 of 69 Official leaderboard
ExploitGym 23.2% #7 of 15 Lab model card
DeepsecBench 19.1 #15 of 29 Official leaderboard
PCT (Polarized Contrastive Pairs) 46.2% #25 of 46 Official leaderboard
BioTIER Permit 99.3% #39 of 69 Official leaderboard
VCT-v2 41.9% #15 of 25 Official leaderboard
First-Person Fairness evaluation 0.88% #6 of 10 Lab model card
ExploitBench 52.9% #13 of 21 Lab model card
Destructive action avoidance 81% #5 of 8 Lab model card
SecIT Bench 71.9% #17 of 26 Official leaderboard
Capture the Flag (Internal) 91.8% #3 of 4 Lab model card
LifeSciBench 56% #4 of 5 Lab model card
MedChemBench (OpenAI) 35% #4 of 5 Lab model card
SEC-bench Pro 57.7% #9 of 11 Lab model card
Suicide Memory Bench 0.53 #10 of 12 Official leaderboard
Instruction hierarchy attacks 99.9% #6 of 7 Lab model card
Prompt injection robustness 96.7% #6 of 7 Lab model card

Other

BenchmarkScoreRankSource
LMCA 55% #18 of 147 Official leaderboard
Conceptual Reasoning Index 65.7 #20 of 151 Official leaderboard
ACCoRD 74.7 #30 of 152 Official leaderboard
Political Science LLM Benchmark 68.3% #9 of 42 Official leaderboard
CAIS Text Capabilities Index 38.2 #23 of 68 Official leaderboard
SAGE v1.1 (Vals) 87% #11 of 29 Official leaderboard
CAIS Risk Index 46.7 #26 of 67 Official leaderboard
SAGE (Vals) 47% #33 of 81 Official leaderboard
Vals Multimodal Index 65.1% #14 of 33 Official leaderboard
CAIS Vision Capabilities Index 55.1 #23 of 53 Official leaderboard
LiveBench 67.2% #16 of 31 Official leaderboard
Artificial Analysis Intelligence Index v4.1 55.0 #6 of 8 Lab model card

Its lab's numbers put it at 158.7, 2.0 points below its ECI

The Self-Reported Index fits one score to the 18 capability results OpenAI reported for this model, on the scale of the Epoch Capabilities Index. Labs choose which benchmarks to show and how to run them, so read it as the lab's claim, not an independent measurement. The 90% interval runs from 155.3 to 162.2. How we compute it.

Its lab's numbers next to 11 leaderboards

Where the benchmark's own leaderboard, or Epoch AI's run, also has this model. Gaps mostly come from different harnesses, effort levels and task subsets. Every pair.

Benchmark Lab says Board says Gap
Agents' Last Exam 50.4% Agents' Last Exam 28% Leaderboard, max, codex +22 pts
OpenAI MRCR v2 89.6% OpenAI MRCR v2 8-needle 256K-512K 89.6 (512K-1M 72.5) 67.9% Leaderboard, 8-needle, 128K to 256K bin, max reasoning +22 pts
Terminal-Bench 2.1 87.4% Terminal-Bench 2.1 78.4% Leaderboard, max, Codex +9.0 pts
AutomationBench 15.2% AutomationBench 23.6% Leaderboard, max −8.4 pts
GeneBench Pro 23.3% GeneBench Pro 28.5% Leaderboard, pro −5.2 pts
FrontierMath Tier 4 v2 68.3% FrontierMath Tier 4 (v2) 70.7% Leaderboard, max; task version 2.0.0, Epoch −2.4 pts
FrontierMath Tiers 1-3 v2 84.9% FrontierMath Tier 1-3 (v2) 86% Leaderboard, max; task version 2.0.0, Epoch −1.1 pts
GPQA Diamond 92.9% GPQA Diamond 93.3% Epoch AI run, max −0.4 pts
DeepSWE 69.6% DeepSWE v1.1 69.6% Leaderboard, max, mini-swe-agent Same
ARC-AGI-3 0.80% ARC-AGI-3 0.80% Leaderboard, max, standard Same
Management Consulting Tasks (OpenAI) 37.2% Management Consulting Tasks (Internal) 37.2% Leaderboard Same

OpenAI's models over time

OpenAI's models in color, against every other model on the Epoch Capabilities Index, by release date.

50 100 150 2023 2024 2025 2026 2027 GPT-6.1 Terra GPT-6 Luna GPT-5.4 Mini GPT-5.2 Pro GPT-5 GPT-5 nano GPT-4.1 nano 50 100 150 2023 2024 2025 2026 2027 GPT-6.1 Terra GPT-6 Luna

Source: Epoch Capabilities Index, Epoch AI (CC BY 4.0)

Its launch post

The benchmarks the lab chose to headline, with the scores it gave.

GPT-5.6 (OpenAI) Jul 9, 2026

Its system cards

Every result in the longer documents the lab published with it. 31 of them aren't in the launch post.

GPT-5.6 Preview System Card Jun 25, 2026

Health

Hallucination

Safety

Robustness

Bias

Monitorability

Bio and chem risk

Cyber risk

AI R&D risk

GPT-5.6 System Card Jul 9, 2026

Health

Hallucination

Safety

Robustness

Bias

Monitorability

Bio and chem risk

Cyber risk

AI R&D risk