AIEvals

Models

GPT-5.6 Terra

Specifications and API prices for GPT-5.6 Terra, recorded by Artificial Analysis.

OpenAIReleased 9 July 2026Context 1M tokensOutput Not published in the checked sourcePrivate weights$2.00 in / $12.00 out per Mtok

Family GPT-5.650 benchmarks with a published resultNo first place on the independent columns among the 46 models this site listsRead 17 October 2026

Where it stands

Rank compares tracked models with a score. Published shows the source’s rank and field size, when available.

BenchmarkScoreRankPublishedLeaderBehind
Vals IndexVals AI53.09%18of 3319of 41Gemini 4 Argon15.81 pts
Artificial Analysis Intelligence IndexArtificial Analysis4220of 42no list readClaude Opus 5.516 pts
Terminal-Bench 4.0Artificial Analysis35.35%13of 40no list readClaude Sonnet 5.528.29 pts
SWE-bench (Verified)Vals AI95.40%5of 235of 88Claude Opus 51.60 pts
LiveCodeBenchVals AI85.93%13of 2126of 143Claude Fable 5.14.59 pts
GPQA DiamondVals AI90.91%17of 2324of 138Gemini 3.1 Pro4.54 pts
MMLU ProVals AI86.66%19of 2440of 138Claude Fable 5.15.72 pts
MMMU ProVals AI86.47%12of 2020of 93Claude Fable 5.14.17 pts
Vibe Code Bench v1.1Vals AI74.59%27of 3830of 106Claude Sonnet 5.517.80 pts
Terminal-Bench Science 0.1Artificial Analysis13.81%10of 24no list readGPT-6 Astra49.52 pts
ProgramBenchVals AI0.50%13of 2913of 54Claude Opus 5.518.00 pts
APEX-Agents-AAArtificial Analysis38.94%3of 7no list readGemini 3 5 Flash8.11 pts
τ³-BankingArtificial Analysis40.21%16of 30no list readQwen3.8 Max11.13 pts
ITBench-AAArtificial Analysis51.04%4of 18no list readGPT-5.6 Sol5.17 pts
IFBenchArtificial Analysis71.22%6of 10no list readGrok 4 20 030911.71 pts
MLCR-AAArtificial Analysis31.67%14of 37no list readClaude Sonnet 5.543.33 pts
ARC-AGI-3ARC Prize · Standard0.80%7of 11no list readGPT-6 Astra61.91 pts
SWE-Bench Pro V2 FullScale AI / SEAL92.37%9of 10no list readClaude Opus 57.03 pts
SWE-Bench Pro V2 HARDScale AI / SEAL86.30%6of 10no list readClaude Opus 511.70 pts
Vibe Code Bench 1–100Vals AI14.82%14of 2014of 20Claude Opus 5.515.54 pts
τ²-Bench TelecomArtificial Analysis86.26%9of 10no list readJt 35b Flash12.86 pts
Harvey LAB-AAArtificial Analysis85.18%18of 20no list readMuse Spark 1.3 Max10.30 pts
EnterpriseOps-Gym-AAArtificial Analysis38.50%17of 22no list readClaude Fable 512.62 pts
AutomationBench-AAArtificial Analysis59.65%16of 40no list readGemini 4 Argon17.86 pts
SciCodeArtificial Analysis54.98%21of 40no list readClaude Opus 5.511.92 pts
GDP.pdfArtificial Analysis24.00%9of 40no list readGPT-6.1 Sol7.00 pts
CritPtArtificial Analysis30.00%7of 42no list readGPT-5.6 Sol2.29 pts
AA-OmniscienceArtificial Analysis0.0537of 42no list readClaude Opus 5.546.37
AA-LCR v1.1Artificial Analysis83.00%12of 42no list readKimi K35.67 pts
Terminal-Bench 2.1Vals AI77.53%14of 3514of 75Claude Opus 5.510.11 pts
ProofBench v1.1Vals AI74.00%14of 3416of 46Claude Sonnet 5.526.00 pts
LegalBenchVals AI85.11%16of 2723of 149Claude Fable 53.45 pts
Harvey's Legal Agent BenchmarkVals AI0.83%35of 3752of 73Muse Spark 1.224.59 pts
SkillsBenchVals AI58.90%9of 2011of 35DeepSeek V4.1 Flash10.90 pts
Vals Multimodal IndexVals AI65.07%12of 1714of 33Claude Fable 59.08 pts
Code MigrationVals AI47.80%11of 3711of 71Claude Sonnet 5.522.03 pts
CyberBench v1.1Vals AI72.08%12of 3114of 43GPT-6 Sol5.90 pts
EMBVals AI66.20%20of 3720of 69Claude Fable 5.110.47 pts
Finance Agent (v2)Vals AI54.44%17of 3722of 73Gemini 4 Argon10.96 pts
IOIVals AI87.61%7of 317of 38Gemini 4 Argon12.39 pts
Legal Research BenchVals AI41.35%20of 3721of 72Muse Spark 1.3 Max13.94 pts
MedCodeVals AI43.41%25of 3543of 104Claude Opus 520.16 pts
MedScribeVals AI82.87%31of 3648of 106Claude Opus 5.58.56 pts
Public Benefits Bench v1.1Vals AI62.38%22of 2924of 45Claude Opus 514.55 pts
SAGEVals AI47.00%19of 3233of 90Claude Opus 4.79.10 pts
Tax Agent BenchVals AI65.20%15of 3717of 64Claude Fable 5.112.44 pts
DeepSWE v1.1Datacurve69.62%6of 206of 28GPT-6 Astra4.50 pts
GDPval-AA v2.1Artificial Analysis1,432.2732of 42no list readClaude Opus 5.5413.9
AA-Briefcase v1.1Artificial Analysis1,333.9529of 41no list readClaude Opus 5.5487.9
Humanity's Last ExamArtificial Analysis42.91%27of 42no list readClaude Opus 5.518.44 pts

Nearby on Vals Index

The two models above and below on the Vals Index (Vals AI), on every column where this model has a published number.

ModelVals Index%Artificial Analysis Intelligence IndexindexTerminal-Bench 4.0%SWE-bench%LiveCodeBench%GPQA Diamond%MMLU Pro%MMMU Pro%Vibe Code Bench v1.1%Terminal-Bench Science 0.1%ProgramBench%
GLM 5.3Z.ai53.51%4541.92%95.40%80.54%88.13%86.77%–78.13%9.52%1.50%
MiMo-V2.6-FlashXiaomi53.23%3822.73%–––––78.96%–0.50%
GPT-5.6 TerraOpenAI53.09%4235.35%95.40%85.93%90.91%86.66%86.47%74.59%13.81%0.50%
Grok 4.6xAI52.09%4417.17%95.60%88.22%94.70%89.40%–76.24%6.19%–
Claude Sonnet 5Anthropic51.77%3814.14%79.60%82.43%88.89%87.55%83.01%81.33%–0.00%
Bold marks a figure better than this model’s on the same column, whichever model it belongs to. Ranks cover the models shown from the stated source.

Cost and speed, as published

Additional measurements from Artificial Analysis’ own row. Read 30 September 2026.

As publishedCost per taskOutput speed
GPT-5.6 Terra (max)$1.4092.7 tok/s

Results by benchmark

Independent and vendor results retain their own harnesses and effort settings.

BenchmarkIndependent runAs reported by a publisher
APEX-Agents-AAKnowledge work38.94%Artificial Analysisnot published
τ³-BankingTool use40.21%Artificial Analysisnot published
ITBench-AAIT operations51.04%Artificial Analysisnot published
IFBenchInstruction following71.22%Artificial Analysisnot published
MLCR-AAMedical reasoning31.67%Artificial Analysisnot published
ARC-AGI-3Interactive reasoning0.80%ARC Prize · Standardnot published
SWE-Bench Pro V2 FullRepository coding92.37%Scale AI / SEALnot published
SWE-Bench Pro V2 HARDRepository coding86.30%Scale AI / SEALnot published
Vibe Code Bench 1–100Application development14.82%Vals AInot published
τ²-Bench TelecomTool use86.26%Artificial Analysisnot published
Harvey LAB-AALegal reasoning85.18%Artificial Analysisnot published
EnterpriseOps-Gym-AAEnterprise workflows38.50%Artificial Analysisnot published
AutomationBench-AAWorkflow automation59.65%Artificial Analysisnot published
SciCodeScientific coding54.98%Artificial Analysisnot published
GDP.pdfDocument reasoning24.00%Artificial Analysisnot published
CritPtScientific reasoning30.00%Artificial Analysisnot published
AA-OmniscienceFactual knowledge0.05Artificial Analysisnot published
AA-LCR v1.1Long-context reasoning83.00%Artificial Analysisnot published
Terminal-Bench 4.0Agentic coding35.35%Artificial Analysis
22.73%Vals AI
not published
Terminal-Bench 2.1Agentic coding77.53%Vals AI
88.01%Artificial Analysis
not published
SWE-bench (Verified)Software engineering95.40%Vals AInot published
LiveCodeBenchCode generation85.93%Vals AInot published
GPQA DiamondScience reasoning90.91%Vals AI
92.53%Artificial Analysis
not published
MMLU ProKnowledge86.66%Vals AInot published
MMMU ProMultimodal reasoning86.47%Vals AI
80.69%Artificial Analysis
not published
Vibe Code Bench v1.1Product coding74.59%Vals AInot published
Terminal-Bench Science 0.1Agentic science13.81%Artificial Analysis
10.00%Vals AI
not published
ProgramBenchAgentic coding0.50%Vals AInot published
ProofBench v1.1Mathematics74.00%Vals AInot published
LegalBenchProfessional knowledge85.11%Vals AInot published
Harvey's Legal Agent BenchmarkProfessional knowledge0.83%Vals AInot published
SkillsBenchAgentic tooling58.90%Vals AInot published
Vals Multimodal IndexComposite65.07%Vals AInot published
Vals IndexComposite53.09%Vals AInot published
Code MigrationAgentic coding47.80%Vals AInot published
CyberBench v1.1Security72.08%Vals AInot published
EMBFinancial modelling66.20%Vals AInot published
Finance Agent (v2)Financial analysis54.44%Vals AInot published
IOICompetitive programming87.61%Vals AInot published
Legal Research BenchProfessional knowledge41.35%Vals AInot published
MedCodeHealthcare administration43.41%Vals AInot published
MedScribeHealthcare administration82.87%Vals AInot published
Public Benefits Bench v1.1Public services62.38%Vals AInot published
SAGEEducation47.00%Vals AInot published
Tax Agent BenchProfessional knowledge65.20%Vals AInot published
DeepSWE v1.1Software engineering69.62%Datacurvenot published
GDPval-AA v2.1Knowledge work1,432.27Artificial Analysisnot published
AA-Briefcase v1.1Knowledge work1,333.95Artificial Analysisnot published
Humanity's Last ExamMultidisciplinary reasoning42.91%Artificial Analysisnot published
Artificial Analysis Intelligence IndexComposite42Artificial Analysisnot published
Independent column: the suite named under each figure, one harness per suite. Publisher column: each lab’s own table, reproduced on the benchmark’s page. Publisher ranks compare reported scores in the same table row only; ties share a rank and unreported cells are excluded.

Sources

Read on 17 October 2026.