BenchLeader
Menu
Best LLMsChoose a modelLeaderboardsModelsBenchmarksProvidersReleases12

GPT-5.6 Terra vs GPT-6 Astra

Full profiles: GPT-5.6 Terra (OpenAI) and GPT-6 Astra (OpenAI): every benchmark, price and rank.

Updated · Data from 10 sources

As of 24 Oct 2026, GPT-6 Astra (max) leads on quality (70.4 vs 64.4 on the BenchLeader Index), but GPT-5.6 Terra (max) costs 4.4× less, so it is the better buy unless you need the extra quality. Pick GPT-5.6 Terra for long context (67 vs 65); pick GPT-6 Astra for knowledge (80 vs 57).

GPT-5.6 Terra benchmarks max OpenAI GPT-6 Astra benchmarks max OpenAIOpen in the comparison tool

Which to pick for…

The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.

Use casePickWhyConfidence
CodingGPT-6 Astra74 vs 65 category score; SciCode 56.5% vs 55.0%High
Agents & tool useGPT-6 Astra64 vs 59 category score; Terminal-Bench 58.2% vs 21.5%Medium
MathsGPT-6 Astra72 vs 69 category score; FrontierMath Tiers 1–3 93.7% vs 86.0%Medium
ReasoningGPT-6 Astra76 vs 70 category score; GPQA Diamond 95.8% vs 93.3%Medium
KnowledgeGPT-6 Astra80 vs 57 category score; SimpleQA Verified 75.6% vs 43.2%High
Long documentsGPT-5.6 Terra67 vs 65 category score; AA-LCR 83.0% vs 80.7%Medium
Images and other mediaGPT-6 Astra66 vs 64 category score; MMMU-Pro 86.9% vs 80.7%Medium
Chat quality (human preference)–Only GPT-6 Astra has an Arena rating (67 category score)Not comparable
Lowest costGPT-5.6 Terra$4.50 vs $20.00 per 1M blended; agentic coding session $0.168 vs $0.800–
Fastest first answerGPT-5.6 Terra91.6 s (GPT-5.6 Terra (max)) vs 247.0 s (GPT-6 Astra (max)) to the first answer token–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • GPT-6 Astra (max) leads on quality: 70.4 vs 64.4.
  • GPT-5.6 Terra (max) is stronger in instruction following, long context.
  • GPT-6 Astra (max) is stronger in agents & tools, coding, composite, knowledge, maths, multimodal, reasoning, human preference.
  • GPT-5.6 Terra (max) is 4.4× cheaper ($4.50 vs $20.00 per 1M blended).
  • GPT-5.6 Terra (max) streams 2.2× faster (126 vs 58 tokens per second).

At a glance

SpecGPT-5.6 TerraGPT-6 Astra
DeveloperOpenAIOpenAI
Released9 Jul 20262 Sept 2026
Context window (tokens)1,050,0001,050,000
Compared atGPT-5.6 Terra (max)GPT-6 Astra (max)
Open weightsNoNo
Price in / out / cached input per 1M$2.00 / $12.00 / $0.200$10.00 / $50.00 / $1.00

Category by category

Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.

CategoryGPT-5.6 Terra (max)GPT-6 Astra (max)Leader
Reasoning70.176.2GPT-6 Astra
Coding65.174.0GPT-6 Astra
Agents & tools58.964.2GPT-6 Astra
Maths68.971.8GPT-6 Astra
Knowledge57.279.5GPT-6 Astra
Instruction following69.9–not comparable
Human preference–66.7not comparable
Multimodal63.766.0GPT-6 Astra
Long context66.665.4GPT-5.6 Terra
Composite76.688.4GPT-6 Astra

Where sources disagree

Across the 56 benchmarks both have, GPT-5.6 Terra leads 7, GPT-6 Astra leads 49. Every overall index that rates both puts GPT-6 Astra ahead.

Overall indexGPT-5.6 Terra (max)GPT-6 Astra (max)Leader
BenchLeader Index64.470.4GPT-6 Astra
AA Intelligence Index v4.3.242.152.7GPT-6 Astra
Arena Text1467†1477GPT-6 Astra

Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).

Price, speed and latency

List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page; suite cost is what Artificial Analysis paid to run its whole Intelligence Index.

MeasureGPT-5.6 Terra (max)GPT-6 Astra (max)
Blended price per 1M tokens$4.50$20.00
Input / output per 1M$2.00 / $12.00$10.00 / $50.00
Cached input per 1M$0.200$1.00
Price per task: chat reply$0.0044$0.019
Price per task: summarize a 30-page report$0.031$0.150
Price per task: code edit$0.030$0.135
Price per task: agentic coding session$0.168$0.800
Price per task: structured extraction$0.0064$0.030
Cost to run a full benchmark suite$1.40$3.26
Output speed126 tok/s58 tok/s
Time to first answer91.6 s247.0 s
Time to a complete answer95.6 s255.6 s

Per token is not per task: on Artificial Analysis' full suite GPT-5.6 Terra (max) cost $1.40 and GPT-6 Astra (max) $3.26, because GPT-5.6 Terra (max) used about 1.9× as many tokens for the same work, which a per-token price hides.

Each model is compared at its best-ranked reasoning-effort setting. GPT-5.6 Terra (max) scores 64.4 and takes 92 s to its first answer token; GPT-5.6 Terra (xhigh) scores 62.8 in 7 s, GPT-5.6 Terra (high) scores 61.3 in 2 s. GPT-6 Astra (max) scores 70.4 and takes 247 s to its first answer token; GPT-6 Astra (high) scores 69.6 in 35 s, GPT-6 Astra (xhigh) scores 68.9 in 130 s. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.

Every benchmark

152 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.

MetricPublisherGPT-5.6 Terra (max)GPT-6 Astra (max)
BenchLeader IndexBenchLeader64.470.4
Agents & tools scoreBenchLeader58.964.2
Coding scoreBenchLeader65.174.0
Composite scoreBenchLeader76.688.4
Instruction following scoreBenchLeader69.9–
Knowledge scoreBenchLeader57.279.5
Long context scoreBenchLeader66.665.4
Maths scoreBenchLeader68.971.8
Multimodal scoreBenchLeader63.766.0
Reasoning scoreBenchLeader70.176.2
Human preference scoreBenchLeader–66.7
Blended price $/MBenchLeader$4.50$20.00
Output speedBenchLeader126 tok/s58 tok/s
Time to first answerBenchLeader91.6 s247.0 s
Context windowBenchLeader1,050,000 tokens1,050,000 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub93.3%95.8%
Humanity's Last ExamScale AI / CAIS–54.8%†
SimpleBenchSimpleBench48.9%†–
Arena Hard PromptsArena (LMArena)1490†1501
LiveBench 2.0 ReasoningLiveBench90.6%†92.7%†
LiveBench 1.0 ReasoningLiveBench90.6%92.7%
GPQA Diamond (AA)Artificial Analysis92.5%96.1%
Humanity's Last Exam (AA)Artificial Analysis42.9%54.7%
GPQA Diamond (Vals)Vals AI90.9%†–
Kagi LLM BenchmarkKagi LLM Benchmark51.3%†–
ARC-AGI-1ARC Prize96.5%97.5%
ARC-AGI-2ARC Prize83.9%95.0%
ARC-AGI-3ARC Prize0.8%98.6%
CritPtArtificial Analysis30.0%31.7%
MysteryMechanismVals AI–53.1%
Chess PuzzlesEpoch AI Benchmarking Hub47.0%74.0%
Mystery Game PuzzlesEpoch AI Benchmarking Hub35.0%84.0%
LMCAEpoch AI Benchmarking Hub55.0%64.1%†
DTBenchEpoch AI Benchmarking Hub93.3%97.1%†
BTF-3Epoch AI Benchmarking Hub–13.5%†
Coding
SciCodeSciCode55.0%56.5%
WeirdMLWeirdML78.3%†93.3%
FrontierCodeCognition41.3%†53.3%
GSO-BenchGSO-Bench–79.4%†
Arena CodingArena (LMArena)1518†1543
Show all 152 benchmarks
MetricPublisherGPT-5.6 Terra (max)GPT-6 Astra (max)
Coding
Arena WebDevArena (LMArena)1521†1775
LiveBench 2.0 CodingLiveBench86.2%†93.7%†
LiveBench 1.0 CodingLiveBench78.3%80.4%
LiveCodeBenchVals AI85.9%†–
IOIVals AI87.6%100.0%
SWE-bench (Vals)Vals AI95.4%–
Code MigrationVals AI47.8%67.7%
Code Migration v1.1Vals AI–80.6%
ProgramBenchVals AI0.5%5.5%
SWE-bench MultimodalVals AI–43.3%
Vibe Code Bench 1-100Vals AI14.8%27.6%
Vibe Code Bench v1.1Vals AI74.6%89.6%
SWE Atlas: Codebase QnAScale AI SEAL–59.1%†
SWE Atlas: RefactoringScale AI SEAL–59.0%†
SWE Atlas: Test WritingScale AI SEAL–50.7%†
MirrorCodeEpoch AI Benchmarking Hub–46.7%†
Surface Evolver BenchEpoch AI Benchmarking Hub83.8%†–
DeepSWE v1.1Epoch AI Benchmarking Hub69.6%73.2%
CursorBenchCursor41.3%–
ALE-BenchEpoch AI Benchmarking Hub1951.42951.3
FrontierSWEEpoch AI Benchmarking Hub–65.5%
Terminal-Bench 4.0 (AA)Artificial Analysis35.4%59.1%
Terminal-Bench 2.1 (AA)Artificial Analysis88.0%88.4%
Startrise FrontendStartrise66.9%†79.9%†
FrontierCode 1.1 (AA)Artificial Analysis–47.0%
Agents & tools
Terminal-BenchTerminal-Bench21.5%58.2%
Remote Labor IndexScale AI / CAIS–20.8%†
APEX-AgentsMercor58.2%64.7%†
Arena AgentArena (LMArena)0.5†14.7
LiveBench Long AgenticLiveBench48.4%†82.4%†
LiveBench BrowserLiveBench84.0%†84.0%†
LiveBench Agentic CodingLiveBench55.0%57.3%
Terminal-Bench HardArtificial Analysis57.6%–
τ²-Bench Telecom (AA)Artificial Analysis86.3%–
Terminal-Bench 2.1 (Vals)Vals AI77.5%87.3%
HiL-BenchScale AI SEAL–65.3%†
GDPval-AA v2.1Artificial Analysis47.6%53.7%
τ³-Banking (AA)Artificial Analysis40.2%41.4%
ITBench SRE (AA)Artificial Analysis51.0%48.6%
Analyst Agent (AA)Artificial Analysis–51.3%
APEX-Agents (AA)Artificial Analysis38.9%–
CUA-benchVals AI–19.2%
CyberBenchVals AI72.1%41.1%
Finance Agent v2Vals AI54.4%53.5%
Finance Agent v2.1Vals AI–60.4%
Harvey's Legal Agent Benchmark v1.1Vals AI–7.1%
Harvey's Legal Agent Benchmark v1Vals AI0.8%5.4%
Legal Research BenchVals AI41.4%39.4%
Robo-benchVals AI–23.3%
SkillsBenchVals AI58.9%–
SREBenchVals AI9.5%56.9%
Tax Agent BenchVals AI30.0%20.7%
Terminal-Bench 4.0 (Vals)Vals AI22.7%59.6%
Terminal-Bench ScienceVals AI10.0%62.9%
Time Horizon Index: KSPVals AI–90.5%
Arena SearchArena (LMArena)–1246
BALROGBALROG53.2%68.3%
Vending-Bench 2Andon Labs7343.2†15514.7†
AutomationBenchArtificial Analysis59.6%68.5%
GDP.pdf (AA)Artificial Analysis24.0%31.0%
Harvey LAB-AA v1.1Harvey–8.6%
EnterpriseOps-GymArtificial Analysis38.5%–
AA-Briefcase v1.1Artificial Analysis–1568
Terminal-Bench-Science 0.1 (AA)Artificial Analysis13.8%63.3%
CWE-Bench (AA)Artificial Analysis–63.3%
DeepSec-Bench (AA)Artificial Analysis–36.9%
CyberGym end-to-end (AA)Artificial Analysis–0.0%
FACTS SearchGoogle81.7%†89.9%†
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub86.0%93.7%
FrontierMath Tier 4Epoch AI Benchmarking Hub70.7%97.6%
OTIS Mock AIMEEpoch AI Benchmarking Hub99.7%100.0%
ProofBenchVals AI74.0%†99.0%†
LiveBench MathematicsLiveBench94.9%96.8%
Arena MathsArena (LMArena)1483†1487
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub43.2%75.6%
LiveBench 2.0 Data AnalysisLiveBench79.3%†83.0%†
LiveBench ResearchLiveBench60.4%†85.2%†
LiveBench 1.0 Data AnalysisLiveBench79.3%83.0%
AA-OmniscienceArtificial Analysis0.143.4
MMLU-ProVals AI86.7%†–
LegalBenchVals AI85.1%–
CorpFinVals AI65.3%†–
TaxEvalVals AI76.2%†–
PRBench FinanceScale AI SEAL–47.5%†
PRBench LegalScale AI SEAL–48.4%†
BioMysteryBenchVals AI–79.3%
Excel Modeling BenchmarkVals AI66.2%71.7%
Excel Modeling Benchmark v1.1Vals AI–72.7%
MedCodeVals AI43.4%†48.5%
MedScribeVals AI82.9%†87.9%
Public Benefits Bench v1.1Vals AI62.4%–
DrugDiscoveryBenchScale AI SEAL–68.7%
EBR-benchEpoch AI Benchmarking Hub–76.2%
GDP.pdfEpoch AI Benchmarking Hub24.7%†34.2%
AA-Omniscience: accuracyArtificial Analysis46.8%62.6%
AA-Omniscience: non-hallucinationArtificial Analysis12.1%48.7%
Instruction following
LiveBench LanguageLiveBench82.9%89.4%
LiveBench Instruction FollowingLiveBench64.6%75.6%
IFBenchArtificial Analysis71.2%–
Arena Instruction FollowingArena (LMArena)1464†1473
Human preference
Arena TextArena (LMArena)1467†1477
EQ-Bench 4EQ-Bench1234†–
Arena Creative WritingArena (LMArena)1429†1450
Arena Multi-turnArena (LMArena)1472†1485
Arena Longer QueriesArena (LMArena)1468†1489
Multimodal
Arena VisionArena (LMArena)1271†1280
LiveBench DesignLiveBench76.0%†82.0%†
MMMU-ProArtificial Analysis80.7%86.9%
MMMU-Pro (Vals)Vals AI86.5%–
MortgageTaxVals AI67.3%†–
SAGEVals AI47.0%46.4%
Blueprint-Bench 2Epoch AI Benchmarking Hub30.8%†49.7%†
AA-3DArtificial Analysis–1166
Long context
AA-LCRArtificial Analysis83.0%80.7%
Arena DocumentArena (LMArena)1477†1473
MLCRArtificial Analysis27.2%41.7%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub160.1†168.8†
LiveBench 2.0LiveBench75.0%†86.1%†
LiveBench 1.0LiveBench77.9%82.2%
AA Intelligence Index v4.3.2Artificial Analysis42.152.7
Vals IndexVals AI–65.0
Vals Finance IndexVals AI–55.6%
Vals RSI IndexVals AI–27.1%
Vals Science IndexVals AI–65.1%
Vals Multimodal IndexVals AI65.1%†–
Vals Router IndexVals AI–65.0%†
AA Cyber IndexArtificial Analysis–33.4

Data as of 2026-10-24. Best configuration of each model; every score links to its source on the model pages.

What changed

  • GPT-5.6 Terra (max): 1 new result (Harvey's Legal Agent Benchmark v1).
  • GPT-6 Astra (max): 9 new results (Code Migration v1.1, Excel Modeling Benchmark v1.1, Finance Agent v2.1…).
  • GPT-5.6 Terra (max): 1 score revision.
  • GPT-6 Astra (max): 1 score revision.
  • GPT-6 Astra (max): 4 score revisions.
  • GPT-5.6 Terra (max): BenchLeader Index 63.8 → 64.4.
  • GPT-6 Astra (max): BenchLeader Index 69.8 → 70.6.
  • GPT-5.6 Terra (max): 5 new results (AutomationBench, GDP.pdf (AA), MLCR…).
  • GPT-5.6 Terra (max): 1 new result (SREBench); 1 score revision.
  • GPT-6 Astra (max): 1 score revision.

Rankings: Best LLM overall · Best LLM for coding · Best LLM for agents · Best cheap LLMs

GPT-5.6 Terra vs GPT-6 Astra: questions

Is GPT-5.6 Terra better than GPT-6 Astra?
GPT-6 Astra (max) leads on quality: 70.4 vs 64.4. The BenchLeader Index combines every independent quality benchmark; GPT-6 Astra (max) is ahead overall as of 2026-10-24, but check the category scores for your use.
Which is cheaper, GPT-5.6 Terra or GPT-6 Astra, and will that change?
GPT-5.6 Terra is cheaper today: $4.50 against $20.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is GPT-5.6 Terra better than GPT-6 Astra for coding?
GPT-6 Astra scores higher in coding (74 vs 65 on the category index, where 50 is average). On SciCode, the top shared coding benchmark, it is 55.0% for GPT-5.6 Terra against 56.5% for GPT-6 Astra.
Is GPT-5.6 Terra better than GPT-6 Astra for agents and tool use?
GPT-6 Astra scores higher in agentic tasks (64 vs 59 on the category index, where 50 is average). On Terminal-Bench, the top shared agents & tools benchmark, it is 21.5% for GPT-5.6 Terra against 58.2% for GPT-6 Astra.
Which is better for long documents?
GPT-5.6 Terra scores higher on long-context benchmarks (67 vs 65). Both accept 1,050,000 tokens of context.
Which has lower latency?
GPT-5.6 Terra (max) answers first: 91.6 s against 247.0 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
Which is faster, GPT-5.6 Terra or GPT-6 Astra?
GPT-5.6 Terra streams faster: 126 against 58 output tokens per second.
Do independent benchmarks agree on GPT-5.6 Terra vs GPT-6 Astra?
Across the 56 benchmarks both have, GPT-5.6 Terra leads 7, GPT-6 Astra leads 49. Every overall index that rates both puts GPT-6 Astra ahead.
When was GPT-6 Astra released?
GPT-6 Astra was released on 2 Sept 2026.