BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases12

GPT-6 Astra vs GPT-6 Sol

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

Updated · Data from 14 sources

As of 9 Nov 2026, GPT-6 Astra is ahead of GPT-6 Sol on quality (68.1 vs 62.9 on the BenchLeader Index), but GPT-6 Sol is 5.0× cheaper at $4.00 vs $20.00 per 1M tokens blended. Choose Astra for reasoning (86 vs 66), coding (83 vs 67), and agents & tools (69 vs 60) and fast first answers (4.8 s vs 72.1 s); choose Sol for knowledge (62 vs 56), where it also streams 3×+ faster.

Key facts (as of 9 Nov 2026)

  • GPT-6 Astra scores 68.1 on the BenchLeader Index and GPT-6 Sol scores 62.9 (BenchLeader, 9 Nov 2026).
  • GPT-6 Astra costs $20.00 per 1M tokens blended (list price, 9 Nov 2026).
  • GPT-6 Sol costs $4.00 per 1M tokens blended (list price, 9 Nov 2026).
  • GPT-6 Sol streams 3×+ faster than GPT-6 Astra (Artificial Analysis, 7-day median, 9 Nov 2026).
  • Of the 3 benchmarks both models have, GPT-6 Astra leads 2 (BenchLeader, 9 Nov 2026).
GPT-6 Astra benchmarks OpenAI GPT-6 Sol benchmarks max OpenAIAdd a third model

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 Astra83 vs 67 category scoreHigh
Agents & tool useGPT-6 Astra69 vs 60 category score; APEX-Agents 64.7% vs 54.3%High
MathsGPT-6 Astra78 vs 69 category scoreHigh
ReasoningGPT-6 Astra86 vs 66 category score; LMCA 64.1% vs 59.1%High
KnowledgeGPT-6 Sol62 vs 56 category scoreMedium
Chat quality (human preference)–Only GPT-6 Sol has an Arena rating (64 category score)Not comparable
Lowest costGPT-6 Sol$4.00 vs $20.00 per 1M blended; agentic coding session $0.160 vs $0.800–
Fastest first answerGPT-6 Astra4.8 s (GPT-6 Astra) vs 72.1 s (GPT-6 Sol (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 leads on quality: 68.1 vs 62.9.
  • GPT-6 Astra is stronger in agents & tools, coding, composite, maths, reasoning.
  • GPT-6 Sol (max) is stronger in knowledge, human preference, long context, multimodal.
  • GPT-6 Sol (max) is 5.0× cheaper ($4.00 vs $20.00 per 1M blended).
  • GPT-6 Sol (max) streams 4.5× faster (121 vs 27 tokens per second).

At a glance

SpecGPT-6 AstraGPT-6 Sol
DeveloperOpenAIOpenAI
Released2 Sep 202622 Sep 2026
Context window (tokens)1,050,0001,050,000
Compared atGPT-6 AstraGPT-6 Sol (max)
Open weightsNoNo
Price in / out / cached input per 1M$10.00 / $50.00 / $1.00$2.00 / $10.00 / $0.200

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-6 AstraGPT-6 Sol (max)Leader
Reasoning85.666.2GPT-6 Astra
Coding83.266.5GPT-6 Astra
Agents & tools69.259.6GPT-6 Astra
Maths77.968.5GPT-6 Astra
Knowledge56.161.5GPT-6 Sol
Human preference–64.2not comparable
Multimodal–60.7not comparable
Long context–66.6not comparable
Composite72.6–not comparable

On one chart

Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.

2550 avg75ReasoningCodingAgents & toolsMathsKnowledgeGPT-6 Astra, Reasoning: 85.6GPT-6 Astra, Coding: 83.2GPT-6 Astra, Agents & tools: 69.2GPT-6 Astra, Maths: 77.9GPT-6 Astra, Knowledge: 56.1GPT-6 Sol, Reasoning: 66.2GPT-6 Sol, Coding: 66.5GPT-6 Sol, Agents & tools: 59.6GPT-6 Sol, Maths: 68.5GPT-6 Sol, Knowledge: 61.5

Not drawn, measured for fewer than two of these models: Instruction following, Long context, Multimodal, Human preference.

ModelIndex

Where sources disagree

Across the 3 benchmarks both have, GPT-6 Astra leads 2, GPT-6 Sol leads 1. Independent sources split: GPT-6 Astra leads on BenchLeader Index and Arena Text; GPT-6 Sol leads on AA Intelligence Index v5.0.

Overall indexGPT-6 AstraGPT-6 Sol (max)Leader
BenchLeader Index68.162.9GPT-6 Astra
AA Intelligence Index v5.043.3†44.2GPT-6 Sol
Arena Text1475†1457GPT-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. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.

MeasureGPT-6 AstraGPT-6 Sol (max)
Blended price per 1M tokens$20.00$4.00
Input / output per 1M$10.00 / $50.00$2.00 / $10.00
Cached input per 1M$1.00$0.200
Price per task: chat reply$0.019$0.0038
Price per task: summarize a 30-page report$0.150$0.030
Price per task: code edit$0.135$0.027
Price per task: agentic coding session$0.800$0.160
Price per task: structured extraction$0.030$0.0060
Agentic coding session, input mostly cached$0.395$0.079
Cost per benchmark task (AA Index)–$2.47
Output tokens per task (AA Index)–54k
Of which reasoning–35k
Time per task (AA Index)–7 min
Output speed27 tok/s121 tok/s
Time to first answer4.8 s72.1 s
Time to a complete answer–130.2 s

Each model is compared at its best-ranked reasoning-effort setting. GPT-6 Astra scores 68.1 and takes 5 s to its first answer token; GPT-6 Astra (max) scores 67.4 in 232 s, GPT-6 Astra (high) scores 66.5 in 64 s. GPT-6 Sol (max) scores 62.9 and takes 72 s to its first answer token; GPT-6 Sol (xhigh) scores 58.6 in 24 s, GPT-6 Sol (high) scores 58.2 in 9 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

144 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-6 AstraGPT-6 Sol (max)
BenchLeader IndexBenchLeader68.162.9
Agents & tools scoreBenchLeader69.259.6
Coding scoreBenchLeader83.266.5
Composite scoreBenchLeader72.6–
Knowledge scoreBenchLeader56.161.5
Maths scoreBenchLeader77.968.5
Reasoning scoreBenchLeader85.666.2
Human preference scoreBenchLeader–64.2
Long context scoreBenchLeader–66.6
Multimodal scoreBenchLeader–60.7
Blended price $/MBenchLeader$20.00$4.00
Output speedBenchLeader27 tok/s121 tok/s
Time to first answerBenchLeader4.8 s72.1 s
Context windowBenchLeader1,050,000 tokens1,050,000 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub95.8%†94.3%
Humanity's Last ExamScale AI / CAIS54.8%–
Arena Hard PromptsArena (LMArena)1500†1483
LiveBench 2.0 ReasoningLiveBench92.7%–
LiveBench 1.0 ReasoningLiveBench92.7%†88.7%
GPQA Diamond (AA)Artificial Analysis96.1%†–
Humanity's Last Exam (AA)Artificial Analysis54.7%†47.9%
HLE-Diamond (AA)Artificial Analysis63.8%†45.2%
ARC-AGI-1ARC Prize97.5%†95.5%
ARC-AGI-2ARC Prize95.0%†89.6%
ARC-AGI-3ARC Prize98.6%†23.0%
CritPtArtificial Analysis31.7%†30.9%
CritPt v2 hard subset (AA)Artificial Analysis70.5%†66.5%
MysteryMechanismVals AI53.1%†30.2%
Chess PuzzlesEpoch AI Benchmarking Hub74.0%†56.0%
Mystery Game PuzzlesEpoch AI Benchmarking Hub84.0%†56.0%
LMCAEpoch AI Benchmarking Hub64.1%59.1%
DTBenchEpoch AI Benchmarking Hub97.1%97.3%
BTF-3Epoch AI Benchmarking Hub13.5%†–
Coding
SciCodeSciCode56.5%†57.6%
WeirdMLWeirdML93.3%†–
FrontierCodeCognition53.3%†49.3%
GSO-BenchGSO-Bench79.4%–
Arena CodingArena (LMArena)1548†1520
Arena WebDevArena (LMArena)1767†1674
Show all 144 benchmarks
MetricPublisherGPT-6 AstraGPT-6 Sol (max)
Coding
LiveBench 2.0 CodingLiveBench87.5%–
LiveBench 1.0 CodingLiveBench80.4%†81.8%
IOIVals AI100.0%†82.6%
Code MigrationVals AI67.7%†57.2%
Code Migration v1.1Vals AI80.6%†–
ProgramBenchVals AI5.5%†2.0%
SWE-bench MultimodalVals AI43.3%†–
Vibe Code Bench 1-100Vals AI27.6%†–
Vibe Code Bench v1.1Vals AI89.6%†87.8%
SWE Atlas: Codebase QnAScale AI SEAL59.1%†–
SWE Atlas: RefactoringScale AI SEAL59.0%†–
SWE Atlas: Test WritingScale AI SEAL50.7%†–
MirrorCodeEpoch AI Benchmarking Hub46.7%†–
DeepSWE v1.1Epoch AI Benchmarking Hub73.2%†–
ALE-BenchEpoch AI Benchmarking Hub2951.3†2462
FrontierSWEEpoch AI Benchmarking Hub65.5%†–
Terminal-Bench 4.0 (AA)Artificial Analysis59.1%†43.9%
Terminal-Bench 2.1 (AA)Artificial Analysis88.4%†–
Startrise FrontendStartrise79.9%–
GameBench 2 OverallDuelLab GameBench 288.772.9†
GameBench 2 by reasoning settingDuelLab GameBench 294.0†79.3†
Design Arena WebsiteDesign Arena1333†1285†
FrontierCode 1.1 (AA)Artificial Analysis47.0%†–
FrontierCode 2.0 (AA)Artificial Analysis42.4%†35.6%
Agents & tools
Terminal-BenchTerminal-Bench60.3%†50.3%
Remote Labor IndexScale AI / CAIS20.8%–
APEX-AgentsMercor64.7%54.3%
Arena AgentArena (LMArena)14.2†7
LiveBench Long AgenticLiveBench75.6%–
LiveBench BrowserLiveBench91.3%–
LiveBench Agentic CodingLiveBench57.3%†52.9%
Terminal-Bench 2.1 (Vals)Vals AI87.3%†83.2%
HiL-BenchScale AI SEAL65.3%–
GDPval-AA v2.1, normalised scoreArtificial Analysis53.7%†–
GDPval-AA v2.1Artificial Analysis1584†–
τ³-Banking (AA)Artificial Analysis41.4%†–
ITBench SRE (AA)Artificial Analysis48.6%†49.4%
Analyst Agent (AA)Artificial Analysis51.3%†–
CUA-benchVals AI19.2%†–
CyberBenchVals AI41.1%†78.0%
Finance Agent v2Vals AI53.5%†49.0%
Finance Agent v2.1Vals AI60.4%†–
Harvey's Legal Agent Benchmark v1.1Vals AI7.1%†–
Harvey's Legal Agent Benchmark v1Vals AI5.4%†1.7%
Legal Research BenchVals AI39.4%†28.9%
Robo-benchVals AI23.3%†–
SREBenchVals AI56.9%†30.5%
Tax Agent BenchVals AI20.7%†15.0%
Terminal-Bench 4.0 (Vals)Vals AI59.6%†44.4%
Terminal-Bench ScienceVals AI62.9%†30.0%
Time Horizon Index: KSPVals AI90.5%†–
Arena SearchArena (LMArena)1245†1229
BALROGBALROG68.3%†–
Vending-Bench 2Andon Labs15514.714427.9†
AutomationBenchArtificial Analysis68.5%†61.6%
GDP.pdf (AA)Artificial Analysis31.0%†25.2%
Harvey LAB-AA v1.1Harvey8.6%†3.6%
AA-Briefcase v1.1Artificial Analysis1569†1476
GDPxGDPx by SemiAnalysis9.6%†–
Design Arena AgenticDesign Arena1393†1212†
Terminal-Bench-Science 0.1 (AA)Artificial Analysis63.3%†30.0%
CWE-Bench (AA)Artificial Analysis63.3%†64.2%
DeepSec-Bench (AA)Artificial Analysis36.9%†46.3%
CyberGym end-to-end (AA)Artificial Analysis0.0%†0.0%
FACTS SearchGoogle93.3%–
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub93.7%†89.8%
FrontierMath Tier 4Epoch AI Benchmarking Hub97.6%†90.0%
OTIS Mock AIMEEpoch AI Benchmarking Hub100.0%†100.0%
ProofBenchVals AI99.0%83.0%†
LiveBench MathematicsLiveBench96.8%†96.4%
Arena MathsArena (LMArena)1481†1440
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub75.6%†60.7%
LiveBench 2.0 Data AnalysisLiveBench83.0%–
LiveBench ResearchLiveBench90.6%–
LiveBench 1.0 Data AnalysisLiveBench83.0%†81.2%
AA-Omniscience v1.1Artificial Analysis42.4†21.4
PRBench FinanceScale AI SEAL47.5%–
PRBench LegalScale AI SEAL48.4%38.6%†
BioMysteryBenchVals AI79.3%†74.8%
Excel Modeling BenchmarkVals AI71.7%†71.5%
Excel Modeling Benchmark v1.1Vals AI72.7%†–
MedCodeVals AI48.5%†47.1%
MedScribeVals AI87.9%†82.0%
Public Benefits Bench v1.1Vals AI–56.6%
DrugDiscoveryBenchScale AI SEAL68.7%†–
EBR-benchEpoch AI Benchmarking Hub76.2%†53.3%
GDP.pdfEpoch AI Benchmarking Hub34.2%†26.4%
AA-Omniscience v1.1: accuracyArtificial Analysis61.6%†51.1%
AA-Omniscience v1.1: non-hallucinationArtificial Analysis50.2%†39.1%
HealthBench Professional (AA)OpenAI66.0%†62.3%
Instruction following
LiveBench LanguageLiveBench89.4%†85.3%
LiveBench Instruction FollowingLiveBench75.6%†68.6%
Arena Instruction FollowingArena (LMArena)1473†1448
Human preference
Arena TextArena (LMArena)1475†1457
Arena Creative WritingArena (LMArena)1450†1430
Arena Multi-turnArena (LMArena)1483†1466
Arena Longer QueriesArena (LMArena)1486†1466
Multimodal
Arena VisionArena (LMArena)1279†1232
LiveBench DesignLiveBench80.0%–
MMMU-ProArtificial Analysis86.9%†83.0%
Blueprint-Bench 2Epoch AI Benchmarking Hub49.7%36.9%†
Code2Video-AAArtificial Analysis1427†–
AA-3DArtificial Analysis1168†1131
Long context
AA-LCR v1.1Artificial Analysis80.7%†83.7%
Arena DocumentArena (LMArena)1473†–
MLCRArtificial Analysis41.7%†33.9%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub169.1164.1†
LiveBench 2.0LiveBench84.5%–
LiveBench 1.0LiveBench82.2%†79.3%
AA Intelligence Index v5.0Artificial Analysis56.9†44.2
AA Intelligence Index v4.3.2Artificial Analysis52.7†47.6
AA Progress Index v1Artificial Analysis349.6†334
Vals IndexVals AI65.0†–
Vals Finance IndexVals AI55.6%†–
Vals RSI IndexVals AI27.1%†28.1%
Vals Science IndexVals AI65.1%†45.0%
Vals Router IndexVals AI65.0%–
AA Cyber IndexArtificial Analysis33.4†36.8
Safety & honesty
FORTRESSScale AI SEAL12.1%13.5%†

Data as of 2026-11-09. Best configuration of each model; every score links to its source on the model pages.

DuelLab GameBench 2, release gb2-20261103-eaea31df94e6 (https://benchmarks.duellab.org/releases/gb2-20261103-eaea31df94e6/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

What changed

  • GPT-6 Sol (max): BenchLeader Index 63.5 → 62.9.
  • GPT-6 Astra: 5 score revisions.
  • GPT-6 Sol (max): 2 score revisions.
  • GPT-6 Sol (max): 1 score revision.
  • GPT-6 Astra: 1 score revision.
  • GPT-6 Astra: 1 new result (FORTRESS); 1 score revision.
  • GPT-6 Sol (max): 1 score revision.
  • GPT-6 Astra: 1 score revision.
  • GPT-6 Sol (max): 6 new results (AA Intelligence Index v5.0, AA-Omniscience v1.1, CritPt v2 hard subset (AA)…).
  • GPT-6 Sol (max): 1 score revision.

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

GPT-6 Astra vs GPT-6 Sol: questions

Is GPT-6 Astra better than GPT-6 Sol?
GPT-6 Astra leads on quality: 68.1 vs 62.9. The BenchLeader Index combines every independent quality benchmark; GPT-6 Astra is ahead overall as of 9 Nov 2026, but check the category scores for your use.
GPT-6 Astra vs GPT-6 Sol: which is cheaper, and will that change?
GPT-6 Sol is cheaper today: $4.00 against $20.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is GPT-6 Astra or GPT-6 Sol better for coding?
GPT-6 Astra scores higher in coding (83 vs 67 on the category index, where 50 is average).
Which is faster, GPT-6 Astra or GPT-6 Sol?
GPT-6 Sol streams 3×+ faster (121 vs 27 output tokens per second, Artificial Analysis 7-day median).
Is GPT-6 Astra or GPT-6 Sol better for agents and tool use?
GPT-6 Astra scores higher in agentic tasks (69 vs 60 on the category index, where 50 is average). On APEX-Agents, the top shared agents & tools benchmark, it is 64.7% for GPT-6 Astra against 54.3% for GPT-6 Sol.
More questions (4)
Which is better for long documents?
Both accept 1,050,000 tokens of context.
Which has lower latency?
GPT-6 Astra answers first: 4.8 s against 72.1 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.
Do independent benchmarks agree on GPT-6 Astra vs GPT-6 Sol?
Across the 3 benchmarks both have, GPT-6 Astra leads 2, GPT-6 Sol leads 1. Independent sources split: GPT-6 Astra leads on BenchLeader Index and Arena Text; GPT-6 Sol leads on AA Intelligence Index v5.0.
When was GPT-6 Sol released?
GPT-6 Sol was released on 22 Sep 2026.