BenchLeader
Menu
Best LLMsChoose a modelLeaderboardsModelsBenchmarksProvidersReleases8

Claude Sonnet 5 vs GPT-6 Astra

Full profiles: Claude Sonnet 5 (Anthropic) and GPT-6 Astra (OpenAI): every benchmark, price and rank.

Updated · Data from 10 sources

As of 29 Oct 2026, GPT-6 Astra is ahead of Claude Sonnet 5 on quality (68.2 vs 59.6 on the BenchLeader Index), but Claude Sonnet 5 is 5.0× cheaper at $4.00 vs $20.00 per 1M tokens blended. GPT-6 Astra leads in every category we measure both in. Claude Sonnet 5 streams ~2× faster.

Key facts (as of 29 Oct 2026)

  • Claude Sonnet 5 scores 59.6 on the BenchLeader Index and GPT-6 Astra scores 68.2 (BenchLeader, 29 Oct 2026).
  • Claude Sonnet 5 costs $4.00 per 1M tokens blended (list price, 29 Oct 2026).
  • GPT-6 Astra costs $20.00 per 1M tokens blended (list price, 29 Oct 2026).
  • Claude Sonnet 5 streams ~2× faster than GPT-6 Astra (Artificial Analysis, 7-day median, 29 Oct 2026).
  • Of the 2 benchmarks both models have, GPT-6 Astra leads 2 (BenchLeader, 29 Oct 2026).
Claude Sonnet 5 benchmarks high Anthropic GPT-6 Astra benchmarks 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 Astra83 vs 61 category scoreHigh
Agents & tool useGPT-6 Astra70 vs 57 category scoreHigh
MathsGPT-6 Astra78 vs 65 category scoreHigh
ReasoningGPT-6 Astra86 vs 66 category score; LMCA 64.1% vs 50.0%High
Chat quality (human preference)–Only Claude Sonnet 5 has an Arena rating (65 category score)Not comparable
Lowest costClaude Sonnet 5$4.00 vs $20.00 per 1M blended; agentic coding session $0.160 vs $0.800–
Fastest first answerEither3.3 s (Claude Sonnet 5 (high)) vs 3.0 s (GPT-6 Astra) 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: 68.2 vs 59.6.
  • Claude Sonnet 5 (high) is stronger in human preference, long context, multimodal.
  • GPT-6 Astra is stronger in agents & tools, coding, maths, reasoning, composite, knowledge.
  • Claude Sonnet 5 (high) is 5.0× cheaper ($4.00 vs $20.00 per 1M blended).
  • Claude Sonnet 5 (high) streams 1.8× faster (71 vs 40 tokens per second).

At a glance

SpecClaude Sonnet 5GPT-6 Astra
DeveloperAnthropicOpenAI
Released30 Jun 20262 Sept 2026
Context window (tokens)1,000,0001,050,000
Compared atClaude Sonnet 5 (high)GPT-6 Astra
Open weightsNoNo
Price in / out / cached input per 1M$2.00 / $10.00 / $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.

CategoryClaude Sonnet 5 (high)GPT-6 AstraLeader
Reasoning65.585.6GPT-6 Astra
Coding60.583.2GPT-6 Astra
Agents & tools57.469.6GPT-6 Astra
Maths65.377.5GPT-6 Astra
Knowledge–55.7not comparable
Human preference64.9–not comparable
Multimodal61.0–not comparable
Long context63.1–not comparable
Composite–73.7not comparable

Where sources disagree

Across the 2 benchmarks both have, Claude Sonnet 5 leads 0, GPT-6 Astra leads 2. Every overall index that rates both puts GPT-6 Astra ahead.

Overall indexClaude Sonnet 5 (high)GPT-6 AstraLeader
BenchLeader Index59.668.2GPT-6 Astra
Arena Text14621477†GPT-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.

MeasureClaude Sonnet 5 (high)GPT-6 Astra
Blended price per 1M tokens$4.00$20.00
Input / output per 1M$2.00 / $10.00$10.00 / $50.00
Cached input per 1M$0.200$1.00
Price per task: chat reply$0.0038$0.019
Price per task: summarize a 30-page report$0.030$0.150
Price per task: code edit$0.027$0.135
Price per task: agentic coding session$0.160$0.800
Price per task: structured extraction$0.0060$0.030
Output speed71 tok/s40 tok/s
Time to first answer3.3 s3.0 s
Time to a complete answer10.3 s–

Each model is compared at its best-ranked reasoning-effort setting. Claude Sonnet 5 (high) scores 59.6 and takes 3 s to its first answer token; Claude Sonnet 5 scores 55.8 in 2 s, Claude Sonnet 5 (xhigh) scores 55.6 in 13 s. GPT-6 Astra scores 68.2 and takes 3 s to its first answer token; GPT-6 Astra (max) scores 67.7 in 232 s, GPT-6 Astra (high) scores 66.8 in 33 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

156 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.

MetricPublisherClaude Sonnet 5 (high)GPT-6 Astra
BenchLeader IndexBenchLeader59.668.2
Agents & tools scoreBenchLeader57.469.6
Coding scoreBenchLeader60.583.2
Human preference scoreBenchLeader64.9–
Long context scoreBenchLeader63.1–
Maths scoreBenchLeader65.377.5
Multimodal scoreBenchLeader61.0–
Reasoning scoreBenchLeader65.585.6
Composite scoreBenchLeader–73.7
Knowledge scoreBenchLeader–55.7
Blended price $/MBenchLeader$4.00$20.00
Output speedBenchLeader71 tok/s40 tok/s
Time to first answerBenchLeader3.3 s3.0 s
Context windowBenchLeader1,000,000 tokens1,050,000 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub90.5%†95.8%†
Humanity's Last ExamScale AI / CAIS–54.8%
SimpleBenchSimpleBench60.6%†–
Arena Hard PromptsArena (LMArena)14901501†
LiveBench 2.0 ReasoningLiveBench–92.7%
LiveBench 1.0 ReasoningLiveBench88.7%†92.7%†
GPQA Diamond (AA)Artificial Analysis91.1%†96.1%†
Humanity's Last Exam (AA)Artificial Analysis35.7%54.7%†
HLE-Diamond (AA)Artificial Analysis–63.8%†
GPQA Diamond (Vals)Vals AI88.9%†–
ARC-AGI-1ARC Prize–97.5%†
ARC-AGI-2ARC Prize–95.0%†
ARC-AGI-3ARC Prize–98.6%†
CritPtArtificial Analysis15.1%31.7%†
CritPt v2 hard subset (AA)Artificial Analysis–70.5%†
MysteryMechanismVals AI–53.1%†
Chess PuzzlesEpoch AI Benchmarking Hub34.0%†74.0%†
Mystery Game PuzzlesEpoch AI Benchmarking Hub16.0%†84.0%†
LMCAEpoch AI Benchmarking Hub50.0%64.1%
DTBenchEpoch AI Benchmarking Hub84.5%97.1%
ForecastBenchForecastBench63.5%†–
BTF-3Epoch AI Benchmarking Hub14.2%†13.5%†
Coding
SciCodeSciCode54.3%56.5%†
WeirdMLWeirdML68.8%93.3%†
FrontierCodeCognition42.7%†53.3%†
Show all 156 benchmarks
MetricPublisherClaude Sonnet 5 (high)GPT-6 Astra
Coding
GSO-BenchGSO-Bench37.3%†79.4%
Arena CodingArena (LMArena)15201543†
Arena WebDevArena (LMArena)15401774†
LiveBench 2.0 CodingLiveBench–93.7%
LiveBench 1.0 CodingLiveBench80.7%†80.4%†
LiveCodeBenchVals AI82.4%†–
IOIVals AI45.0%†100.0%†
SWE-bench (Vals)Vals AI79.6%†–
Code MigrationVals AI44.4%†67.7%†
Code Migration v1.1Vals AI–80.6%†
ProgramBenchVals AI0.0%†5.5%†
SWE-bench MultimodalVals AI–43.3%†
Vibe Code Bench 1-100Vals AI13.8%†27.6%†
Vibe Code Bench v1.1Vals AI81.3%†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 Hub60.0%†–
DeepSWE v1.1Epoch AI Benchmarking Hub48.2%73.2%†
CursorBenchCursor30.8%–
ALE-BenchEpoch AI Benchmarking Hub1463.12951.3†
FrontierSWEEpoch AI Benchmarking Hub–65.5%†
Terminal-Bench 4.0 (AA)Artificial Analysis5.0%59.1%†
Terminal-Bench 2.1 (AA)Artificial Analysis80.5%†88.4%†
Startrise FrontendStartrise66.6%†79.9%
FrontierCode 1.1 (AA)Artificial Analysis–47.0%†
FrontierCode 2.0 (AA)Artificial Analysis–42.4%†
Agents & tools
Terminal-BenchTerminal-Bench12.4%†58.2%†
Remote Labor IndexScale AI / CAIS–20.8%
APEX-AgentsMercor54.5%†64.7%
Arena AgentArena (LMArena)3.814†
LiveBench Long AgenticLiveBench–77.5%
LiveBench BrowserLiveBench–91.3%
LiveBench Agentic CodingLiveBench59.4%†57.3%†
Terminal-Bench 2.1 (Vals)Vals AI74.5%†87.3%†
HiL-BenchScale AI SEAL–65.3%
GDPval-AA v2.1, normalised scoreArtificial Analysis–53.7%†
GDPval-AA v2.1Artificial Analysis–1584†
τ³-Banking (AA)Artificial Analysis37.3%†41.4%†
ITBench SRE (AA)Artificial Analysis–48.6%†
Analyst Agent (AA)Artificial Analysis46.3%†51.3%†
CUA-benchVals AI–19.2%†
CyberBenchVals AI61.9%†41.1%†
Finance Agent v2Vals AI53.9%†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 AI5.0%†5.4%†
Legal Research BenchVals AI41.8%†39.4%†
Robo-benchVals AI–23.3%†
SkillsBenchVals AI46.5%†–
SREBenchVals AI0.4%†56.9%†
Tax Agent BenchVals AI29.4%†20.7%†
Terminal-Bench 4.0 (Vals)Vals AI10.6%†59.6%†
Terminal-Bench ScienceVals AI5.7%†62.9%†
Time Horizon Index: KSPVals AI–90.5%†
Arena SearchArena (LMArena)–1246†
BALROGBALROG–68.3%†
GBAEvalEpoch AI Benchmarking Hub65.3%†–
Vending-Bench 2Andon Labs6377.7†15514.7
AutomationBenchArtificial Analysis32.1%68.5%†
GDP.pdf (AA)Artificial Analysis9.4%31.0%†
Harvey LAB-AA v1.1Harvey–8.6%†
EnterpriseOps-GymArtificial Analysis43.9%†–
AA-Briefcase v1.1Artificial Analysis1356†1568†
Terminal-Bench-Science 0.1 (AA)Artificial Analysis–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 SearchGoogle–89.9%
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub65.6%†93.7%†
FrontierMath Tier 4Epoch AI Benchmarking Hub29.3%†97.6%†
OTIS Mock AIMEEpoch AI Benchmarking Hub94.7%†100.0%†
ProofBenchVals AI77.0%†99.0%
LiveBench MathematicsLiveBench92.9%†96.8%†
Arena MathsArena (LMArena)14731487†
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub32.9%†75.6%†
LiveBench 2.0 Data AnalysisLiveBench–83.0%
LiveBench ResearchLiveBench–85.2%
LiveBench 1.0 Data AnalysisLiveBench71.7%†83.0%†
AA-Omniscience v1.1Artificial Analysis–42.4†
MMLU-ProVals AI87.5%†–
LegalBenchVals AI83.9%†–
CorpFinVals AI67.0%†–
TaxEvalVals AI75.6%†–
PRBench FinanceScale AI SEAL–47.5%
PRBench LegalScale AI SEAL–48.4%
BioMysteryBenchVals AI–79.3%†
Excel Modeling BenchmarkVals AI66.3%†71.7%†
Excel Modeling Benchmark v1.1Vals AI–72.7%†
MedCodeVals AI47.5%†48.5%†
MedScribeVals AI76.0%†87.9%†
Public Benefits Bench v1.1Vals AI66.0%†–
DrugDiscoveryBenchScale AI SEAL50.0%†68.7%†
EBR-benchEpoch AI Benchmarking Hub–76.2%†
GDP.pdfEpoch AI Benchmarking Hub–34.2%†
AA-Omniscience v1.1: accuracyArtificial Analysis–61.6%†
AA-Omniscience v1.1: non-hallucinationArtificial Analysis–50.2%†
HealthBench Professional (AA)OpenAI–66.0%†
Instruction following
LiveBench LanguageLiveBench75.0%†89.4%†
LiveBench Instruction FollowingLiveBench63.9%†75.6%†
Arena Instruction FollowingArena (LMArena)14661473†
Human preference
Arena TextArena (LMArena)14621477†
EQ-Bench 4EQ-Bench1221†–
Arena Creative WritingArena (LMArena)14351450†
Arena Multi-turnArena (LMArena)14731485†
Arena Longer QueriesArena (LMArena)14811489†
Multimodal
Arena VisionArena (LMArena)12731280†
LiveBench DesignLiveBench–82.0%
MMMU-ProArtificial Analysis77.3%†86.9%†
MMMU-Pro (Vals)Vals AI83.0%†–
MortgageTaxVals AI70.0%†–
Blueprint-Bench 2Epoch AI Benchmarking Hub24.9%†49.7%
Code2Video-AAArtificial Analysis–1427†
AA-3DArtificial Analysis–1164†
Long context
AA-LCR v1.1Artificial Analysis76.7%80.7%†
Arena DocumentArena (LMArena)14761473†
MLCRArtificial Analysis4.4%41.7%†
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub156.9†168.8
LiveBench 2.0LiveBench–85.6%
LiveBench 1.0LiveBench76.0%†82.2%†
AA Intelligence Index v5.0Artificial Analysis–56.9†
AA Intelligence Index v4.3.2Artificial Analysis31.752.7†
AA Progress Index v1Artificial Analysis296.8349.6†
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 AI68.8%†–
Vals Router IndexVals AI–65.0%
AA Cyber IndexArtificial Analysis–33.4†

Data as of 2026-10-29. Best configuration of each model; every score links to its source on the model pages.

What changed

  • Claude Sonnet 5 (high): 1 new result (AA Progress Index v1).
  • GPT-6 Astra: 3 score revisions.
  • GPT-6 Astra: 1 new result (Vals Router Index).
  • Claude Sonnet 5 (high): BenchLeader Index 60.0 → 59.5.
  • Claude Sonnet 5 (high): 1 new result (MLCR).
  • GPT-6 Astra: 1 score revision.
  • Claude Sonnet 5 (high): 1 score revision.
  • GPT-6 Astra: 8 new results (LiveBench 2.0, LiveBench Long Agentic, LiveBench 2.0 Reasoning…).
  • Claude Sonnet 5 (high): 2 new results (AutomationBench, GDP.pdf (AA)).
  • GPT-6 Astra: 1 new result (Startrise Frontend).

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

Claude Sonnet 5 vs GPT-6 Astra: questions

Is GPT-6 Astra better than Claude Sonnet 5?
GPT-6 Astra leads on quality: 68.2 vs 59.6. The BenchLeader Index combines every independent quality benchmark; GPT-6 Astra is ahead overall as of 29 Oct 2026, but check the category scores for your use.
Claude Sonnet 5 vs GPT-6 Astra: which is cheaper, and will that change?
Claude Sonnet 5 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 Claude Sonnet 5 or GPT-6 Astra better for coding?
GPT-6 Astra scores higher in coding (83 vs 61 on the category index, where 50 is average).
Which is faster, Claude Sonnet 5 or GPT-6 Astra?
Claude Sonnet 5 streams ~2× faster (71 vs 40 output tokens per second, Artificial Analysis 7-day median).
Is Claude Sonnet 5 or GPT-6 Astra better for agents and tool use?
GPT-6 Astra scores higher in agentic tasks (70 vs 57 on the category index, where 50 is average).
More questions (4)
Which is better for long documents?
Both take about 1.1M tokens: Claude Sonnet 5 1,000,000, GPT-6 Astra 1,050,000.
Which has lower latency?
GPT-6 Astra answers first: 3.0 s against 3.3 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 Claude Sonnet 5 vs GPT-6 Astra?
Across the 2 benchmarks both have, Claude Sonnet 5 leads 0, GPT-6 Astra leads 2. 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.