BenchLeader

GPT-6.2 Astra vs Kimi K3.1 Pro

Full profiles: GPT-6.2 Astra (OpenAI) and Kimi K3.1 Pro (Moonshot AI): every benchmark, price and rank.

Updated · Data from 13 sources

As of 19 Nov 2026, GPT-6.2 Astra is ahead of Kimi K3.1 Pro on quality (68.9 vs 60.7 on the BenchLeader Index), but Kimi K3.1 Pro is 4.0× cheaper at $4.00 vs $16.00 per 1M tokens blended. GPT-6.2 Astra leads in every category we measure both in. GPT-6.2 Astra streams ~1.5× faster.

Key facts (as of 19 Nov 2026)

  • GPT-6.2 Astra scores 68.9 on the BenchLeader Index and Kimi K3.1 Pro scores 60.7 (BenchLeader, 19 Nov 2026).
  • GPT-6.2 Astra costs $16.00 per 1M tokens blended (list price, 19 Nov 2026).
  • Kimi K3.1 Pro costs $4.00 per 1M tokens blended (list price, 19 Nov 2026).
  • GPT-6.2 Astra streams ~1.5× faster than Kimi K3.1 Pro (Artificial Analysis, 7-day median, 19 Nov 2026).
  • Of the 22 benchmarks both models have, GPT-6.2 Astra leads 16 (BenchLeader, 19 Nov 2026).

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-6.2 Astra76 vs 59 category score; FrontierCode 2.0 (AA) 45.4% vs 25.4%High
Agents & tool useGPT-6.2 Astra66 vs 60 category score; CWE-Bench (AA) 65.0% vs 60.8%Medium
MathsEither73 vs 73 category scoreEither
ReasoningGPT-6.2 Astra73 vs 58 category score; HLE-Diamond (AA) 72.2% vs 37.3%High
KnowledgeGPT-6.2 Astra74 vs 57 category score; AA-Omniscience v1.1 50 vs 29.9High
Images and other mediaGPT-6.2 Astra72 vs 66 category score; MMMU-Pro 89.5% vs 83.9%Medium
Lowest costKimi K3.1 Pro$4.00 vs $16.00 per 1M blended; agentic coding session $0.160 vs $0.640–
Fastest first answerKimi K3.1 Pro155.6 s (Kimi K3.1 Pro) vs 202.7 s (GPT-6.2 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-6.2 Astra (max) leads on quality: 68.9 vs 60.7.
  • GPT-6.2 Astra (max) is stronger in agents & tools, coding, knowledge, multimodal, reasoning.
  • Kimi K3.1 Pro is stronger in maths.
  • Kimi K3.1 Pro is 4.0× cheaper ($4.00 vs $16.00 per 1M blended).
  • GPT-6.2 Astra (max) streams 1.6× faster (49 vs 31 tokens per second).

At a glance

SpecGPT-6.2 AstraKimi K3.1 Pro
DeveloperOpenAIMoonshot AI
Released17 Nov 202623 Oct 2026
Context window (tokens)1,050,0001,048,576
Compared atGPT-6.2 Astra (max)Kimi K3.1 Pro
Open weightsNoYes
Price in / out / cached input per 1M$8.00 / $40.00 / $0.400$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-6.2 Astra (max)Kimi K3.1 ProLeader
Reasoning73.157.9GPT-6.2 Astra
Coding76.158.8GPT-6.2 Astra
Agents & tools66.459.5GPT-6.2 Astra
Maths72.573.4level
Knowledge74.257.1GPT-6.2 Astra
Multimodal71.766.0GPT-6.2 Astra
Long context–67.9not comparable
Composite69.5–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 & toolsMathsKnowledgeMultimodalGPT-6.2 Astra, Reasoning: 73.1GPT-6.2 Astra, Coding: 76.1GPT-6.2 Astra, Agents & tools: 66.4GPT-6.2 Astra, Maths: 72.5GPT-6.2 Astra, Knowledge: 74.2GPT-6.2 Astra, Multimodal: 71.7Kimi K3.1 Pro, Reasoning: 57.9Kimi K3.1 Pro, Coding: 58.8Kimi K3.1 Pro, Agents & tools: 59.5Kimi K3.1 Pro, Maths: 73.4Kimi K3.1 Pro, Knowledge: 57.1Kimi K3.1 Pro, Multimodal: 66.0

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

Model, click to add or removeIndex

Where sources disagree

Across the 22 benchmarks both have, GPT-6.2 Astra leads 16, Kimi K3.1 Pro leads 6. Every overall index that rates both puts GPT-6.2 Astra ahead.

Overall indexGPT-6.2 Astra (max)Kimi K3.1 ProLeader
BenchLeader Index68.960.7GPT-6.2 Astra
AA Intelligence Index v5.0.162.146.5GPT-6.2 Astra
Vals Index67.762.5†GPT-6.2 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-6.2 Astra (max)Kimi K3.1 Pro
Blended price per 1M tokens$16.00$4.00
Input / output per 1M$8.00 / $40.00$2.00 / $10.00
Cached input per 1M$0.400$0.200
Price per task: chat reply$0.015$0.0038
Price per task: summarize a 30-page report$0.120$0.030
Price per task: code edit$0.108$0.027
Price per task: agentic coding session$0.640$0.160
Price per task: structured extraction$0.024$0.0060
Agentic coding session, input mostly cached$0.298$0.079
Cost per benchmark task (AA Index)$4.00$3.68
Output tokens per task (AA Index)41k72k
Of which reasoning26k44k
Time per task (AA Index)14 min39 min
Output speed49 tok/s31 tok/s
Time to first answer202.7 s155.6 s
Time to a complete answer, last measured 13 Nov 2026–417.5 s

Per token is not per task: an average task of Artificial Analysis' Intelligence Index cost $4.00 on GPT-6.2 Astra (max) and $3.68 on Kimi K3.1 Pro, and Kimi K3.1 Pro wrote 1.8× as many output tokens for the same work, which a per-token price hides.

Each model is compared at its best-ranked reasoning-effort setting. GPT-6.2 Astra (max) scores 68.9 and takes 203 s to its first answer token; GPT-6.2 Astra (xhigh) scores 67.1 in 67 s, GPT-6.2 Astra (medium) scores 65.6 in 4 s. Kimi K3.1 Pro scores 60.7 and takes 156 s to its first answer token; Kimi K3.1 Pro (max) scores 60.2 in 156 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

118 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-6.2 Astra (max)Kimi K3.1 Pro
BenchLeader IndexBenchLeader68.960.7
Agents & tools scoreBenchLeader66.459.5
Coding scoreBenchLeader76.158.8
Composite scoreBenchLeader69.5–
Knowledge scoreBenchLeader74.257.1
Maths scoreBenchLeader72.573.4
Multimodal scoreBenchLeader71.766.0
Reasoning scoreBenchLeader73.157.9
Long context scoreBenchLeader–67.9
Blended price $/MBenchLeader$16.00$4.00
Output speedBenchLeader49 tok/s31 tok/s
Time to first answerBenchLeader202.7 s155.6 s
Context windowBenchLeader1,050,000 tokens1,048,576 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub96.8%–
Arena Hard PromptsArena (LMArena)–1505†
LiveBench 2.0 ReasoningLiveBench91.6%91.1%†
Humanity's Last Exam (AA)Artificial Analysis–52.5%
HLE-Diamond (AA)Artificial Analysis72.2%37.3%
ARC-AGI-1ARC Prize98.5%–
ARC-AGI-2ARC Prize95.8%–
ARC-AGI-3ARC Prize99.8%–
CritPtArtificial Analysis–24.6%
CritPt v2 hard subset (AA)Artificial Analysis71.5%49.5%
MysteryMechanismVals AI53.1%26.1%†
HLE-Diamond (Scale)Scale AI SEAL66.9%†–
Arena ExpertArena (LMArena)–1516†
Chess PuzzlesEpoch AI Benchmarking Hub78.0%–
Mystery Game PuzzlesEpoch AI Benchmarking Hub95.0%–
LMCAEpoch AI Benchmarking Hub–57.6%†
DTBenchEpoch AI Benchmarking Hub–93.3%†
Coding
SciCodeSciCode–60.8%
Arena CodingArena (LMArena)–1537†
Arena WebDevArena (LMArena)17811685†
LiveBench 2.0 CodingLiveBench91.9%67.8%†
IOIVals AI–81.9%†
Vibe Code Bench v1.1Vals AI89.9%84.9%†
Vibe Code Bench 1-100 v1.1Vals AI55.2%–
SWE Atlas v1.1Scale AI SEAL60.5%†61.1%
Show all 118 benchmarks
MetricPublisherGPT-6.2 Astra (max)Kimi K3.1 Pro
Coding
SWE Atlas v1.1: Codebase QnAScale AI SEAL67.7%†70.2%
SWE Atlas v1.1: Test WritingScale AI SEAL62.2%†58.9%
SWE Atlas v1.1: RefactoringScale AI SEAL51.4%†54.3%
MirrorCodeEpoch AI Benchmarking Hub67.8%–
ALE-BenchEpoch AI Benchmarking Hub–1932.8†
WeirdML v3WeirdML–9.9%†
Terminal-Bench 4.0 (AA)Artificial Analysis64.1%67.2%
GameBench 2 OverallDuelLab GameBench 2–78.3
GameBench 2 by reasoning settingDuelLab GameBench 2–84.0†
Design Arena WebsiteDesign Arena1339†1318
FrontierCode 1.1 (AA)Artificial Analysis–47.1%
FrontierCode 2.0 (AA)Artificial Analysis45.4%25.4%
Agents & tools
Terminal-BenchTerminal-Bench62.1%59.7%†
APEX-AgentsMercor–65.5%
Arena AgentArena (LMArena)15.23.9†
LiveBench Long AgenticLiveBench70.8%38.7%†
LiveBench BrowserLiveBench93.0%75.9%†
HiL-BenchScale AI SEAL–61.3%
GDPval-AA v2.1, normalised scoreArtificial Analysis–67.3%
GDPval-AA v2.1Artificial Analysis–1861
Analyst Agent (AA)Artificial Analysis62.5%38.8%
Finance Agent v2.1Vals AI60.8%61.5%†
Harvey's Legal Agent Benchmark v1.1Vals AI16.3%11.7%†
Legal Research BenchVals AI42.8%37.5%†
Robo-benchVals AI32.7%0.4%†
SREBenchVals AI–27.1%†
Tax Agent BenchVals AI21.8%24.9%†
Terminal-Bench 4.0 (Vals)Vals AI63.1%63.6%†
Terminal-Bench ScienceVals AI71.4%31.4%†
Arena Agent Task CompletionArena (LMArena)12.41.9†
SpireClimbEpoch AI Benchmarking Hub41.3%–
AutomationBenchArtificial Analysis77.3%73.3%
GDP.pdf (AA)Artificial Analysis29.8%31.4%
Harvey LAB-AA v1.1Harvey15.0%–
EnterpriseOps-GymArtificial Analysis29.9%–
AA-Briefcase v1.1Artificial Analysis16431674
Design Arena AgenticDesign Arena13041291
Terminal-Bench-Science 0.1 (AA)Artificial Analysis78.1%34.3%
CWE-Bench (AA)Artificial Analysis65.0%60.8%
DeepSec-Bench (AA)Artificial Analysis46.8%36.7%
CyberGym end-to-end (AA)Artificial Analysis0.0%94.7%
MKA (AA)Artificial Analysis52.7%43.9%
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub92.6%–
FrontierMath Tier 4Epoch AI Benchmarking Hub100.0%–
OTIS Mock AIMEEpoch AI Benchmarking Hub100.0%–
ProofBenchVals AI–91.0%
ArXivMathMathArena–58.3%†
BrokenArXivMathArena–61.9%†
Autoform-HardVals AI87.5%–
Arena MathsArena (LMArena)–1485†
FrontierMath ErdősEpoch AI Benchmarking Hub10.3%–
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub78.5%–
LiveBench 2.0 Data AnalysisLiveBench84.1%79.1%†
LiveBench ResearchLiveBench91.5%82.1%†
AA-Omniscience v1.1Artificial Analysis5029.9
PRBench FinanceScale AI SEAL–47.5%
PRBench LegalScale AI SEAL43.1%†47.9%
BioMysteryBenchVals AI80.7%75.6%†
Excel Modeling Benchmark v1.1Vals AI76.0%70.0%†
Access to Justice BenchmarkVals AI–69.5%†
DrugDiscoveryBenchScale AI SEAL65.5%†63.4%†
EBR-benchEpoch AI Benchmarking Hub92.4%–
GDP.pdfEpoch AI Benchmarking Hub–46.6%†
AA-Omniscience v1.1: accuracyArtificial Analysis68.6%48.9%
AA-Omniscience v1.1: non-hallucinationArtificial Analysis40.9%62.7%
Instruction following
MultiChallengeScale AI SEAL73.2%†–
Arena Instruction FollowingArena (LMArena)–1493†
Human preference
Arena TextArena (LMArena)–1487†
EQ-Bench Creative Writing v3EQ-Bench–2173
Arena Creative WritingArena (LMArena)–1456†
Arena Multi-turnArena (LMArena)–1477†
Arena Longer QueriesArena (LMArena)–1489†
Multimodal
LiveBench DesignLiveBench82.8%73.7%†
MMMU-ProArtificial Analysis89.5%83.9%
Humanity's Sixth SenseScale AI SEAL63.2%†33.2%†
Furniture AssemblyEpoch AI Benchmarking Hub89.2%–
VIALS-AAHandshake AI65.8%45.3%
AA-3DArtificial Analysis13181112
Long context
AA-LCR v1.1Artificial Analysis–86.3%
MLCRArtificial Analysis39.4%38.3%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub174.2†161.4
LiveBench 2.0LiveBench84.9%68.7%†
AA Intelligence Index v5.0.1Artificial Analysis62.146.5
AA Intelligence Index v4.3.2Artificial Analysis–55.1
AA Progress Index v1Artificial Analysis–340.7
Vals IndexVals AI67.762.5†
Vals Finance IndexVals AI57.3%56.0%†
Vals RSI IndexVals AI29.0%–
Vals Science IndexVals AI68.4%44.4%†
Vals Legal IndexVals AI69.7%67.0%†
AA Cyber IndexArtificial Analysis37.364.1
Safety & honesty
FORTRESSScale AI SEAL–18.6%
Metadata
AA Openness IndexArtificial Analysis–33.3

Data as of 2026-11-19. Best configuration of each model; every score links to its source on the model pages.

DuelLab GameBench 2, release gb2-20261114-ca113f475f5e (https://benchmarks.duellab.org/releases/gb2-20261114-ca113f475f5e/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

What changed

  • GPT-6.2 Astra (max): BenchLeader Index 67.8 → 68.9.
  • Kimi K3.1 Pro: BenchLeader Index 61.2 → 60.7.
  • GPT-6.2 Astra (max): 6 new results (VIALS-AA, Arena WebDev, Arena Agent…).
  • GPT-6.2 Astra (max): BenchLeader Index 67.2 → 67.8.
  • GPT-6.2 Astra (max): 2 new results (MKA (AA), Terminal-Bench).
  • Kimi K3.1 Pro: BenchLeader Index 61.8 → 61.2.
  • Kimi K3.1 Pro: 1 score revision.
  • Kimi K3.1 Pro: BenchLeader Index 61.2 → 61.8.
  • Kimi K3.1 Pro: 1 score revision.
  • Kimi K3.1 Pro: 1 score revision.

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

GPT-6.2 Astra vs Kimi K3.1 Pro: questions

Is GPT-6.2 Astra better than Kimi K3.1 Pro?
GPT-6.2 Astra (max) leads on quality: 68.9 vs 60.7. The BenchLeader Index combines every independent quality benchmark; GPT-6.2 Astra (max) is ahead overall as of 19 Nov 2026, but check the category scores for your use.
GPT-6.2 Astra vs Kimi K3.1 Pro: which is cheaper, and will that change?
Kimi K3.1 Pro is cheaper today: $4.00 against $16.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is GPT-6.2 Astra or Kimi K3.1 Pro better for coding?
GPT-6.2 Astra scores higher in coding (76 vs 59 on the category index, where 50 is average). On FrontierCode 2.0 (AA), the top shared coding benchmark, it is 45.4% for GPT-6.2 Astra against 25.4% for Kimi K3.1 Pro.
Which is faster, GPT-6.2 Astra or Kimi K3.1 Pro?
GPT-6.2 Astra streams ~1.5× faster (49 vs 31 output tokens per second, Artificial Analysis 7-day median).
Is GPT-6.2 Astra or Kimi K3.1 Pro better for agents and tool use?
GPT-6.2 Astra scores higher in agentic tasks (66 vs 60 on the category index, where 50 is average). On CWE-Bench (AA), the top shared agents & tools benchmark, it is 65.0% for GPT-6.2 Astra against 60.8% for Kimi K3.1 Pro.
More questions (4)
Which is better for long documents?
Both take about 1.1M tokens: GPT-6.2 Astra 1,050,000, Kimi K3.1 Pro 1,048,576.
Which has lower latency?
Kimi K3.1 Pro answers first: 155.6 s against 202.7 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-6.2 Astra vs Kimi K3.1 Pro?
Across the 22 benchmarks both have, GPT-6.2 Astra leads 16, Kimi K3.1 Pro leads 6. Every overall index that rates both puts GPT-6.2 Astra ahead.
When was GPT-6.2 Astra released?
GPT-6.2 Astra was released on 17 Nov 2026.