Skip to main content

OpenAIMoonshot AILLM matchup

GPT-6.2 Astravs.Kimi K3

Which one is better?

GPT-6.2 Astra performs better than Kimi K3, but costs 2.7 times as much.

GPT-6.2 Astra ahead
IntelligenceOverall index, 100 points = average
126.6

GPT-6.2 Astra

Rank 4

101.7

Kimi K3

Rank 29

SpeedTokens per second
n/a

GPT-6.2 Astra

59

Kimi K3

Kimi K3 cheaper
Price$ per 1M tokens
$16

GPT-6.2 Astra

$6

Kimi K3

Popularity0 to 100
n/a

GPT-6.2 Astra

86

Kimi K3

Rank 4 of 511

With more than 272,000 input tokens, GPT-6.2 Astra costs more: $16 input and $60 output per 1M tokens.

Choose GPT-6.2 Astra for

CodingReasoningAgentsGeneralLegalFinance (not certain)

No clear advantage for Kimi K3.

Where is the difference?

Strengths at a glance

GPT-6.2 AstraKimi K3Median of all ranked models

11013015090CodingGPT-6.2 AstraReasoningGPT-6.2 AstraAgentsGPT-6.2 AstraGeneralGPT-6.2 AstraFinanceslightlyGPT-6.2 AstraLegalGPT-6.2 Astra
CodingGPT-6.2 Astra ahead
GPT-6.2 Astra131.1Rank 1
Kimi K396.7Rank 46
Median of ranked models84.3

95 % range: GPT-6.2 Astra 121.7 to 140.5, Kimi K3 88.4 to 104.9.

AreaGPT-6.2 AstraKimi K3

The same values and verdicts as the bars below. In the close-up the net starts at 90 points, just below the weaker model's lowest value; the reference group average is 100. Not in the net because a model has no own value there or the data are too thin: Mathematics, Knowledge, Multimodal, Health.

GPT-6.2 Astra is ahead in 5 of 10 areas, Kimi K3 in none.

GPT-6.2 Astra ahead: 5GPT-6.2 Astra slightly ahead: 1too little data: 4

Areas in the overall index
CodingGPT-6.2 Astra +34131.1 to 96.7
ReasoningGPT-6.2 Astra +27124.2 to 97.6
AgentsGPT-6.2 Astra +25122.2 to 97.7
GeneralGPT-6.2 Astra +23128.9 to 106.4
Knowledgetoo little data133.1 to 95.4
Multimodaltoo little data129.7 to 101.6
Mathematicstoo little data128.4 to 101.5
Specialist fields
LegalGPT-6.2 Astra +9118.8 to 110.0
Financeleaning GPT-6.2 Astranot certain · 3 and 8 tests
Healthonly Kimi K3 has a score

Bar length shows the gap in index points (full width 15 points, one standard deviation of the reference group). Solid means clearly ahead, hatched not certain yet, a thin grey bar level.

Show 3 more areas that do not count toward the overall index
Cybersecurityonly Kimi K3 has a score
Productivityonly Kimi K3 has a score
Educationonly Kimi K3 has a score

Bar length shows the gap in index points (full width 15 points, one standard deviation of the reference group). Solid means clearly ahead, hatched not certain yet, a thin grey bar level.

GPT-6.2 Astra vs. Kimi K3 at the same effort

Both models measured at the same reasoning efforts. This shows whether a lead comes from the model or only from more compute.

lowGPT-6.2 Astra ahead
34.4 points gap across 10 shared test series
Overall index (max vs. max)GPT-6.2 Astra ahead
GPT-6.2 Astra 126.6 vs. Kimi K3 101.7

The gap compares only the Artificial Analysis test series in which both models are measured at that effort. Tied below 3 points, ahead only beyond two standard errors.

Which model for which task?

Everyday tasks, translated from the area values above. Each row names the values it rests on.

Writing and debugging codeGPT-6.2 AstraCoding 131.1 to 96.7
Maths and data analysisGPT-6.2 AstraMathematics 128.4 to 101.5 (too little data), Reasoning 124.2 to 97.6
Facts and researchopenKnowledge 133.1 to 95.4 (too little data)
Agents and workflowsGPT-6.2 AstraAgents 122.2 to 97.7
Security analysisopenCybersecurity without a score for GPT-6.2 Astra
Many requests on a small budgetKimi K3$16 to $6 per 1M tokens
Finance, legal and medical textsGPT-6.2 AstraFinance 111.3 to 105.8, Legal 118.8 to 110.0, Health without a score for GPT-6.2 Astra
Images, screenshots and PDFsopenMultimodal 129.7 to 101.6 (too little data)

GPT-6.2 Astra vs. Kimi K3 in our own test

The same prompt for both models. We show the latest attempt without fixes; it does not count toward the overall index.

Pelican on a bicycle

Prompt

Generate an SVG of a pelican riding a bicycle

Task origin: Simon Willison

GPT-6.2 Astra

Not tested yet

Kimi K3

Kimi K3: Pelican on a bicycle
A community classic for free-form SVG drawing.

Oct 1, 2026 · 173 s · $0.08

Test details
Run date
Oct 1, 2026
API provider
openrouter
Requested API model
moonshotai/kimi-k3
Reasoning setting
high
Token limit including reasoning
8,192 tokens
Output tokens
5,145 tokens
Cost of the attempt
$0.0775
Response time
173.4 s
API model reported in response
moonshotai/kimi-k3
Pinned endpoint
Moonshot AI | moonshotai/kimi-k3-20260715
Temperature
Provider default
Test protocol
community-visual-tests-v1
Technical checks
Passed, not a quality rating

Visually reviewed. This individual test does not establish overall model quality.

Earlier attempts (1)

  • Sep 7, 2026API request failedReasoning setting: high
SVG short test

There is no shared test round for this selection. Results from different rounds are not mixed.

GPT-6.2 Astra and Kimi K3 in the overall index

Every model with a place in the overall index, both models of this matchup highlighted. You can add or hide other models.

Overall index (Gradually Meta-Score v6)

20 of 99 models

Claude Fable 5.2max130.3
Gemini 4 Carbonhigh127.8
Gemini 4 Neonxhigh127.1
GPT-6.2 Astramax126.6
Claude Opus 5.5max125.0
Claude Fable 5.1max124.1
Claude Sonnet 5.5max121.9
Gemini 4 Argonhigh121.5
GPT-6 Astramax119.6
GPT-6.2 Solmax119.2
Muse Wonder Fullmax117.9
GPT-6.1 Solmax117.3
Claude Opus 5max117.1
Muse Wondermax113.5
Claude Fable 5max112.8
Kimi K3.1 Promax112.4
Muse Ember 1.0max111.1
GPT-5.6 Solmax109.5
GPT-6 Solmax108.3
Kimi K3max101.7

Popularity compared

How often the models are searched for, used and mentioned, on a scale from 0 to 100.

GPT-6.2 Astra has no popularity score yet.

Over time

The timeline appears once both models have two values.

All data side by side

The better value is bold. For the overall index only when the gap is clear.

MetricGPT-6.2 AstraKimi K3
Intelligence
Overall indexpoints, 100 = reference group average, higher is better126.6101.7
RankRank 4 of 99Rank 29 of 99
Price per 1M tokens
Inputwhat you send$8above 272,000 input tokens: $16$3
Outputwhat the model writes$40above 272,000 input tokens: $60$15
Speed
Tokens per secondn/a59
Popularity
Popularitysearch, usage and mentions, 0 to 100n/a86Rank 4 of 511
Technical
Context windowsource-reported context limit1.05 M tokens≈ 1,580 A4 pages1.05 M tokens≈ 1,570 A4 pages
Knowledge cutoffMay 18, 2026n/a
ReleasedNov 17, 2026Jul 16, 2026
AvailabilityActiveActive
Open weightsnoyes
  • GPT-6.2 Astra, 12-hour cache write price: The table lists the 30-minute rate. For 12-hour retention, OpenAI charges $20 per 1M tokens at the Standard rate through 272,000 input tokens. Source
  • GPT-6.2 Astra, Long-context pricing: Above 272,000 input tokens, the higher rates apply to the entire request. Source
  • GPT-6.2 Astra, Batch, Fast, and Ultrafast: Batch costs half the Standard rate and Fast costs twice the Standard rate. Ultrafast costs three times the Standard rate through February 28, 2027 and six times the rate from March 1, 2027. The change follows Pacific time. Source

Prices without caching or batch discounts · Page count as a rule of thumb from Artificial Analysis

Is there a better choice?

Successor of Kimi K3Kimi K3.1 Pro

Overall index 112.4, rank 16$2 input and $10 output per 1M tokens

View profile
Cheaper than Kimi K3Gemini 4 Carbon

Overall index 127.8, rank 2$2 input and $10 output per 1M tokens

GPT-6.2 Astra vs. Gemini 4 Carbon
Best model in the overall indexClaude Fable 5.2

Overall index 130.3, rank 1$8 input and $40 output per 1M tokens

GPT-6.2 Astra vs. Claude Fable 5.2

Shared benchmarks

Single results from the same benchmark version, task and metric. A value is bold only with the same test setup and a clear gap.

A: GPT-6.2 AstraB: Kimi K3

AA-AnalystAgentOverall · higher is betterGPT-6.2 Astra: 62.5%Kimi K3: 38.8%Artificial AnalysisAs of 11/20/2026
AA-BriefcaseOverall · higher is betterGPT-6.2 Astra: 1,038.6 eloKimi K3: 985.1 eloArtificial AnalysisAs of 11/20/2026
AA-OmniscienceOverall · higher is betterGPT-6.2 Astra: 38.667 pointsKimi K3: -1.700 pointsArtificial AnalysisAs of 11/20/2026
AA-Omniscience AccuracyOverall · higher is betterGPT-6.2 Astra: 60.6%Kimi K3: 42.4%Artificial AnalysisAs of 11/20/2026
AA-Omniscience Hallucination RateOverall · lower is betterGPT-6.2 Astra: 55.8%Kimi K3: 76.6%Artificial AnalysisAs of 11/20/2026
AutomationBench-AAOverall · higher is betterGPT-6.2 Astra: 68.6%Kimi K3: 49.3%Artificial AnalysisAs of 11/20/2026
CritPt v2 (Hard Subset)Overall · higher is betterGPT-6.2 Astra: 61.0%Kimi K3: 5.5%Artificial AnalysisAs of 11/20/2026
FrontierCodeOverall · higher is betterGPT-6.2 Astra: 32.7%Kimi K3: 10.9%Artificial AnalysisAs of 11/20/2026
GDP.pdfOverall · higher is betterGPT-6.2 Astra: 28.0%Kimi K3: 18.6%Artificial AnalysisAs of 11/20/2026
HLE-DiamondOverall · higher is betterGPT-6.2 Astra: 47.7%Kimi K3: 14.0%Artificial AnalysisAs of 11/20/2026
Show 3 more benchmarks
MMMU-ProOverall · higher is betterGPT-6.2 Astra: 89.5%Kimi K3: 80.5%Artificial AnalysisAs of 11/20/2026
Terminal-BenchOverall · higher is betterGPT-6.2 Astra: 51.0%Kimi K3: 12.6%Artificial AnalysisAs of 11/20/2026
Terminal-Bench-ScienceOverall · higher is betterGPT-6.2 Astra: 51.0%Kimi K3: 1.0%Artificial AnalysisAs of 11/20/2026

More comparisons

Compare GPT-6.2 Astra and Kimi K3 with the models people compare them with most often.

Compare another matchup