BenchLeader
Menu
Best LLMsChoose a modelLeaderboardsModelsBenchmarksProvidersReleases8

GPT-6 Astra vs Kimi K2.5

Full profiles: GPT-6 Astra (OpenAI) and Kimi K2.5 (Moonshot AI): every benchmark, price and rank.

Updated · Data from 13 sources

As of 29 Oct 2026, GPT-6 Astra is ahead of Kimi K2.5 on quality (68.2 vs 58.8 on the BenchLeader Index), but Kimi K2.5 is 17× cheaper at $1.20 vs $20.00 per 1M tokens blended. Choose GPT-6 Astra for coding (83 vs 56), agents & tools (70 vs 45), and maths (78 vs 56) and fast first answers (3.0 s vs 36.0 s); choose Kimi K2.5 for knowledge (60 vs 56), where it also streams ~2× faster.

Key facts (as of 29 Oct 2026)

  • GPT-6 Astra scores 68.2 on the BenchLeader Index and Kimi K2.5 scores 58.8 (BenchLeader, 29 Oct 2026).
  • GPT-6 Astra costs $20.00 per 1M tokens blended (list price, 29 Oct 2026).
  • Kimi K2.5 costs $1.20 per 1M tokens blended (list price, 29 Oct 2026).
  • Kimi K2.5 streams ~2× faster than GPT-6 Astra (Artificial Analysis, 7-day median, 29 Oct 2026).
GPT-6 Astra benchmarks OpenAI Kimi K2.5 benchmarks thinking Moonshot AIOpen 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 56 category scoreHigh
Agents & tool useGPT-6 Astra70 vs 45 category scoreHigh
MathsGPT-6 Astra78 vs 56 category scoreHigh
ReasoningGPT-6 Astra86 vs 67 category scoreHigh
KnowledgeKimi K2.560 vs 56 category scoreMedium
Chat quality (human preference)–Only Kimi K2.5 has an Arena rating (63 category score)Not comparable
Lowest costKimi K2.5$1.20 vs $20.00 per 1M blended; agentic coding session $0.048 vs $0.800–
Fastest first answerGPT-6 Astra3.0 s (GPT-6 Astra) vs 36.0 s (Kimi K2.5 (thinking)) 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 58.8.
  • GPT-6 Astra is stronger in agents & tools, coding, composite, maths, reasoning.
  • Kimi K2.5 (thinking) is stronger in knowledge, human preference, instruction following, long context, multimodal.
  • Kimi K2.5 (thinking) is 17× cheaper ($1.20 vs $20.00 per 1M blended).
  • Kimi K2.5 (thinking) streams 2.2× faster (87 vs 40 tokens per second).

At a glance

SpecGPT-6 AstraKimi K2.5
DeveloperOpenAIMoonshot AI
Released2 Sept 202627 Jan 2026
Context window (tokens)1,050,000262,144
Compared atGPT-6 AstraKimi K2.5 (thinking)
Open weightsNoYes
Price in / out / cached input per 1M$10.00 / $50.00 / $1.00$0.600 / $3.00 / $0.100

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 AstraKimi K2.5 (thinking)Leader
Reasoning85.667.3GPT-6 Astra
Coding83.255.7GPT-6 Astra
Agents & tools69.645.1GPT-6 Astra
Maths77.556.2GPT-6 Astra
Knowledge55.760.0Kimi K2.5
Instruction following–69.0not comparable
Human preference–63.4not comparable
Multimodal–59.3not comparable
Long context–63.8not comparable
Composite73.7–not comparable

Where sources disagree

Across the 0 benchmarks both have, GPT-6 Astra leads 0, Kimi K2.5 leads 0. Every overall index that rates both puts GPT-6 Astra ahead.

Overall indexGPT-6 AstraKimi K2.5 (thinking)Leader
BenchLeader Index68.258.8GPT-6 Astra
Arena Text1477†1450GPT-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-6 AstraKimi K2.5 (thinking)
Blended price per 1M tokens$20.00$1.20
Input / output per 1M$10.00 / $50.00$0.600 / $3.00
Cached input per 1M$1.00$0.100
Price per task: chat reply$0.019$0.0011
Price per task: summarize a 30-page report$0.150$0.0090
Price per task: code edit$0.135$0.0081
Price per task: agentic coding session$0.800$0.048
Price per task: structured extraction$0.030$0.0018
Output speed40 tok/s87 tok/s
Time to first answer3.0 s36.0 s
Time to a complete answer–41.7 s

Each model is compared at its best-ranked reasoning-effort setting. 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. Kimi K2.5 (thinking) scores 58.8 and takes 36 s to its first answer token; Kimi K2.5 scores 54.3 in 1 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

172 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 AstraKimi K2.5 (thinking)
BenchLeader IndexBenchLeader68.258.8
Agents & tools scoreBenchLeader69.645.1
Coding scoreBenchLeader83.255.7
Composite scoreBenchLeader73.7–
Knowledge scoreBenchLeader55.760.0
Maths scoreBenchLeader77.556.2
Reasoning scoreBenchLeader85.667.3
Human preference scoreBenchLeader–63.4
Instruction following scoreBenchLeader–69.0
Long context scoreBenchLeader–63.8
Multimodal scoreBenchLeader–59.3
Blended price $/MBenchLeader$20.00$1.20
Output speedBenchLeader40 tok/s87 tok/s
Time to first answerBenchLeader3.0 s36.0 s
Context windowBenchLeader1,050,000 tokens262,144 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub95.8%†87.6%†
Humanity's Last ExamScale AI / CAIS54.8%24.4%†
SimpleBenchSimpleBench–46.8%†
Arena Hard PromptsArena (LMArena)1501†1471
LiveBench 2.0 ReasoningLiveBench92.7%–
LiveBench 1.0 ReasoningLiveBench92.7%†–
GPQA Diamond (AA)Artificial Analysis96.1%†87.9%
Humanity's Last Exam (AA)Artificial Analysis54.7%†30.7%
HLE-Diamond (AA)Artificial Analysis63.8%†–
GPQA Diamond (Vals)Vals AI–84.1%
Kagi LLM BenchmarkKagi LLM Benchmark–78.5%
ARC-AGI-1ARC Prize97.5%†65.3%†
ARC-AGI-2ARC Prize95.0%†11.8%†
ARC-AGI-3ARC Prize98.6%†–
CritPtArtificial Analysis31.7%†3.1%
CritPt v2 hard subset (AA)Artificial Analysis70.5%†–
MysteryMechanismVals AI53.1%†–
Chess PuzzlesEpoch AI Benchmarking Hub74.0%†–
Mystery Game PuzzlesEpoch AI Benchmarking Hub84.0%†–
CL-benchEpoch AI Benchmarking Hub–19.3%†
CL-bench LifeEpoch AI Benchmarking Hub–13.2%†
LMCAEpoch AI Benchmarking Hub64.1%–
DTBenchEpoch AI Benchmarking Hub97.1%–
BTF-3Epoch AI Benchmarking Hub13.5%†–
Coding
SWE-bench Verified (Epoch)Epoch AI Benchmarking Hub–73.8%†
Show all 172 benchmarks
MetricPublisherGPT-6 AstraKimi K2.5 (thinking)
Coding
SciCodeSciCode56.5%†49.0%†
WeirdMLWeirdML93.3%†45.6%†
FrontierCodeCognition53.3%†–
GSO-BenchGSO-Bench79.4%–
Arena CodingArena (LMArena)1543†1502
Arena WebDevArena (LMArena)1774†1437
LiveBench 2.0 CodingLiveBench93.7%–
LiveBench 1.0 CodingLiveBench80.4%†–
SciCode (AA)Artificial Analysis56.5%†–
LiveCodeBenchVals AI–83.9%
IOIVals AI100.0%†–
SWE-bench (Vals)Vals AI–70.0%
SWE-bench Verified (bash only)SWE-bench–70.8%†
SWE-bench Verified (any scaffold)SWE-bench–70.8%†
Code MigrationVals AI67.7%†7.0%
Code Migration v1.1Vals AI80.6%†–
ProgramBenchVals AI5.5%†–
SWE-bench MultimodalVals AI43.3%†–
Vibe Code Bench 1-100Vals AI27.6%†–
Vibe Code Bench v1.1Vals AI89.6%†17.5%
SWE Atlas: Codebase QnAScale AI SEAL59.1%†13.1%†
SWE Atlas: RefactoringScale AI SEAL59.0%†20.9%†
SWE Atlas: Test WritingScale AI SEAL50.7%†25.8%†
MirrorCodeEpoch AI Benchmarking Hub46.7%†–
DeepSWE v1.1Epoch AI Benchmarking Hub73.2%†–
ALE-BenchEpoch AI Benchmarking Hub2951.3†821.6†
FrontierSWEEpoch AI Benchmarking Hub65.5%†–
Terminal-Bench 4.0 (AA)Artificial Analysis59.1%†–
Terminal-Bench 2.1 (AA)Artificial Analysis88.4%†45.7%
Startrise FrontendStartrise79.9%–
FrontierCode 1.1 (AA)Artificial Analysis47.0%†–
FrontierCode 2.0 (AA)Artificial Analysis42.4%†–
Agents & tools
Terminal-BenchTerminal-Bench58.2%†43.2%†
Remote Labor IndexScale AI / CAIS20.8%–
APEX-AgentsMercor64.7%–
Arena AgentArena (LMArena)14†–
LiveBench Long AgenticLiveBench77.5%–
LiveBench BrowserLiveBench91.3%–
LiveBench Agentic CodingLiveBench57.3%†–
Terminal-Bench HardArtificial Analysis–34.9%
τ²-Bench Telecom (AA)Artificial Analysis–95.9%
Terminal-Bench 2.1 (Vals)Vals AI87.3%†42.0%
MCP AtlasScale AI SEAL–64.4%†
HiL-BenchScale AI SEAL65.3%–
GDPval-AA v2.1, normalised scoreArtificial Analysis53.7%†–
GDPval-AA v2.1Artificial Analysis1584†–
τ³-Banking (AA)Artificial Analysis41.4%†14.2%
ITBench SRE (AA)Artificial Analysis48.6%†–
Analyst Agent (AA)Artificial Analysis51.3%†–
APEX-Agents (AA)Artificial Analysis–11.5%
CUA-benchVals AI19.2%†–
CyberBenchVals AI41.1%†–
Finance Agent v2Vals AI53.5%†35.8%
Finance Agent v2.1Vals AI60.4%†–
Harvey's Legal Agent Benchmark v1.1Vals AI7.1%†–
Harvey's Legal Agent Benchmark v1Vals AI5.4%†0.0%
Legal Research BenchVals AI39.4%†15.9%
Robo-benchVals AI23.3%†–
SREBenchVals AI56.9%†–
Tax Agent BenchVals AI20.7%†–
Terminal-Bench 2.0 (Vals)Vals AI–40.5%
Terminal-Bench 4.0 (Vals)Vals AI59.6%†–
Terminal-Bench ScienceVals AI62.9%†–
Time Horizon Index: KSPVals AI90.5%†–
Arena SearchArena (LMArena)1246†–
BALROGBALROG68.3%†–
Vending-Bench 2Andon Labs15514.71198.5†
AutomationBenchArtificial Analysis68.5%†–
GDP.pdf (AA)Artificial Analysis31.0%†–
Harvey LAB-AA v1.1Harvey8.6%†–
AA-Briefcase v1.1Artificial Analysis1568†–
Terminal-Bench-Science 0.1 (AA)Artificial Analysis63.3%†–
CWE-Bench (AA)Artificial Analysis63.3%†–
DeepSec-Bench (AA)Artificial Analysis36.9%†–
CyberGym end-to-end (AA)Artificial Analysis0.0%†–
FACTS SearchGoogle89.9%–
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub93.7%†–
FrontierMath Tier 4Epoch AI Benchmarking Hub97.6%†–
OTIS Mock AIMEEpoch AI Benchmarking Hub100.0%†92.2%†
ProofBenchVals AI99.0%–
LiveBench MathematicsLiveBench96.8%†–
AIME (Vals)Vals AI–95.6%
AIME 2026MathArena–95.8%
HMMT February 2026MathArena–87.1%
MathArena ApexMathArena–8.8%
Arena MathsArena (LMArena)1487†1471
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub75.6%†34.3%†
LiveBench 2.0 Data AnalysisLiveBench83.0%–
LiveBench ResearchLiveBench85.2%–
LiveBench 1.0 Data AnalysisLiveBench83.0%†–
AA-Omniscience v1.1Artificial Analysis42.4†–
MMLU-ProVals AI–85.9%
CorpFinVals AI–68.3%
TaxEvalVals AI–74.2%
MedQAVals AI–94.4%
PRBench FinanceScale AI SEAL47.5%46.5%†
PRBench LegalScale AI SEAL48.4%43.8%†
MultiNRCScale AI SEAL–35.2%†
BioMysteryBenchVals AI79.3%†–
CaseLaw v2Vals AI–58.7%
Excel Modeling BenchmarkVals AI71.7%†28.5%
Excel Modeling Benchmark v1.1Vals AI72.7%†–
MedCodeVals AI48.5%†39.3%
MedScribeVals AI87.9%†76.4%
DrugDiscoveryBenchScale AI SEAL68.7%†–
EBR-benchEpoch AI Benchmarking Hub76.2%†–
GDP.pdfEpoch AI Benchmarking Hub34.2%†–
AA-Omniscience v1.1: accuracyArtificial Analysis61.6%†–
AA-Omniscience v1.1: non-hallucinationArtificial Analysis50.2%†–
HealthBench Professional (AA)OpenAI66.0%†–
Instruction following
LiveBench LanguageLiveBench89.4%†–
LiveBench Instruction FollowingLiveBench75.6%†–
IFBenchArtificial Analysis–70.2%
MultiChallengeScale AI SEAL–61.4%†
TutorBenchScale AI SEAL–54.6%†
Arena Instruction FollowingArena (LMArena)1473†1439
Human preference
Arena TextArena (LMArena)1477†1450
Arena Creative WritingArena (LMArena)1450†1423
Arena Multi-turnArena (LMArena)1485†1452
Arena Longer QueriesArena (LMArena)1489†1458
Multimodal
Arena VisionArena (LMArena)1280†1269
LiveBench DesignLiveBench82.0%–
MMMU-ProArtificial Analysis86.9%†75.4%
VISTAScale AI SEAL–41.9%†
MMMU-Pro (Vals)Vals AI–84.3%
MortgageTaxVals AI–66.5%
Blueprint-Bench 2Epoch AI Benchmarking Hub49.7%–
Code2Video-AAArtificial Analysis1427†–
AA-3DArtificial Analysis1164†–
Long context
Fiction.LiveBench 120kFiction.live–78.1%†
AA-LCR v1.1Artificial Analysis80.7%†78.0%
Arena DocumentArena (LMArena)1473†1436
MLCRArtificial Analysis41.7%†–
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub168.8148.1†
LiveBench 2.0LiveBench85.6%–
LiveBench 1.0LiveBench82.2%†–
AA Intelligence Index v5.0Artificial Analysis56.9†–
AA Intelligence Index v4.3.2Artificial Analysis52.7†23.5
AA Progress Index v1Artificial Analysis349.6†270.5
Vals IndexVals AI65.0†–
Vals Finance IndexVals AI55.6%†–
Vals RSI IndexVals AI27.1%†6.8%†
Vals Science IndexVals AI65.1%†–
Vals Router IndexVals AI65.0%–
AA Cyber IndexArtificial Analysis33.4†–
Safety & honesty
FORTRESSScale AI SEAL–41.1%†
MASKScale AI SEAL–70.5%†

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

What changed

  • Kimi K2.5 (thinking): 1 new result (AA Progress Index v1).
  • GPT-6 Astra: 3 score revisions.
  • GPT-6 Astra: 1 new result (Vals Router Index).
  • Kimi K2.5 (thinking): 1 new result (Harvey's Legal Agent Benchmark v1).
  • GPT-6 Astra: 1 score revision.
  • Kimi K2.5 (thinking): output price moved from $2.75/M to $3.00/M (9% up).
  • GPT-6 Astra: 8 new results (LiveBench 2.0, LiveBench Long Agentic, LiveBench 2.0 Reasoning…).
  • Kimi K2.5 (thinking): BenchLeader Index 59.4 → 58.9.
  • GPT-6 Astra: 1 new result (Startrise Frontend).
  • GPT-6 Astra: BenchLeader Index 68.9 → 68.0.

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

GPT-6 Astra vs Kimi K2.5: questions

Is GPT-6 Astra better than Kimi K2.5?
GPT-6 Astra leads on quality: 68.2 vs 58.8. 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.
GPT-6 Astra vs Kimi K2.5: which is cheaper, and will that change?
Kimi K2.5 is cheaper today: $1.20 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 Kimi K2.5 better for coding?
GPT-6 Astra scores higher in coding (83 vs 56 on the category index, where 50 is average).
Which is faster, GPT-6 Astra or Kimi K2.5?
Kimi K2.5 streams ~2× faster (87 vs 40 output tokens per second, Artificial Analysis 7-day median).
Is GPT-6 Astra or Kimi K2.5 better for agents and tool use?
GPT-6 Astra scores higher in agentic tasks (70 vs 45 on the category index, where 50 is average).
More questions (3)
Which is better for long documents?
GPT-6 Astra accepts more context: 1,050,000 against 262,144 tokens for Kimi K2.5.
Which has lower latency?
GPT-6 Astra answers first: 3.0 s against 36.0 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.
When was GPT-6 Astra released?
GPT-6 Astra was released on 2 Sept 2026.