Head-to-head comparison

Grok 4.6 vs. Kimi K3

Compare published benchmark results from matching versions and cohorts, API pricing, and technical specifications. Any documented test setup differences remain visible.

shared result series: 46
distinct benchmarks: 24
latest retrieval: 2026-09-04

First modelSecond model

Specifications

General specifications

Release date, availability, context window, architecture, and published parameter counts at a glance. Undisclosed details remain marked as unavailable.

Specifications for Grok 4.6 and Kimi K3
Attribute	Grok 4.6	Kimi K3
Provider	xAI	Moonshot AI
Released	Aug 12, 2026	Jul 16, 2026
Availability	Active	Active
Model type	Proprietary	Open Weights
Parameters	Not published	2.8T, 104B active
Architecture	Not published	Mixture of Experts
Context window	500,000 tokens	1,048,576 tokens
Knowledge cutoff	February 1, 2026	Not published
Technical sources	Model	Model

Pricing

API pricing compared

The table compares official standard API rates per 1M tokens, including published cache pricing. Batch, fast, priority, regional surcharges, tool calls, and cloud platform pricing are excluded.

API pricing for Grok 4.6 and Kimi K3
Price type	Grok 4.6	Kimi K3
API input	$2 ≤200K / $4 >200K	$3
API output	$6 ≤200K / $12 >200K	$15
Cache read	$0.5 ≤200K / $1 >200K	$0.3
Cache write (5 min.)	Not listed	Not listed
Cache write (1 hr.)	Not listed	Not listed
Price verified	08/16/2026 Source	07/30/2026 Source

Grok 4.6Kimi K3

API inputUSD per 1M tokens, base tier

Grok 4.6$2

Kimi K3$3

API outputUSD per 1M tokens, base tier

Grok 4.6$6

Kimi K3$15

Cache readUSD per 1M tokens, base tier

Grok 4.6$0.5

Kimi K3$0.3

Key to the marks:better valuebetter, but not conclusivetie

Pricing by provider

Direct developer pricing and provider endpoint tariffs routed through OpenRouter are listed separately. Prices are in USD per 1M tokens. Context-dependent direct tiers plus regional, Flex, Priority, and other OpenRouter tariffs remain individually identifiable.

Provider pricing for Grok 4.6 and Kimi K3
Model	Provider	Purchase route and tariff	Input per 1M tokens	Output per 1M tokens	Source
Grok 4.6	xAI	Direct from developerUp to 200,000 context tokens	Lowest listed price: $2	Lowest listed price: $6	Source08/16/2026
	xAI	Direct from developerAbove 200,000 context tokens	$4	$12	Source08/16/2026
	Amazon Bedrock	Via OpenRouteramazon-bedrock/us-west-2	$2.2	$6.6	Source09/06/2026
	xAI	Via OpenRouterxai	Lowest listed price: $2	Lowest listed price: $6	Source09/06/2026
	xAI	Via OpenRouterxai/priority	$4	$12	Source09/06/2026
	xAI	Via OpenRouterxai/zdr	Lowest listed price: $2	Lowest listed price: $6	Source09/06/2026
	xAI	Via OpenRouterxai/zdr/priority	$4	$12	Source09/06/2026
Kimi K3	Moonshot AI	Direct from developerStandard API	$3	$15	Source07/30/2026
	Alibaba	Via OpenRouteralibaba	$3.45	$17.25	Source09/06/2026
	BaseTen	Via OpenRouterbaseten/fp8	$3	$15	Source09/06/2026
	Chutes	Via OpenRouterchutes/mxfp4	$3	$15	Source09/06/2026
	DeepInfra	Via OpenRouterdeepinfra/bf16	$2.85	$14.25	Source09/06/2026
	DigitalOcean	Via OpenRouterdigitalocean	$2.85	$14.25	Source09/06/2026
	Fireworks	Via OpenRouterfireworks	$3	$15	Source09/06/2026
	Fireworks	Via OpenRouterfireworks/fast	$4.5	$22.5	Source09/06/2026
	Fireworks	Via OpenRouterfireworks/us	$3.3	$16.5	Source09/06/2026
	Makora	Via OpenRoutermakora	Lowest listed price: $2.55	Lowest listed price: $12.75	Source09/06/2026
	Modal	Via OpenRoutermodal/mxfp4	$3	$15	Source09/06/2026
	Moonshot AI	Via OpenRoutermoonshotai/mxfp4	$3	$15	Source09/06/2026
	Morph	Via OpenRoutermorph/fast	$6	$22.5	Source09/06/2026
	Morph	Via OpenRoutermorph/fp4	$2.6	$14	Source09/06/2026
	Parasail	Via OpenRouterparasail/fp4	$3	$15	Source09/06/2026
	Phala	Via OpenRouterphala	$3	$15	Source09/06/2026
	Sail Research	Via OpenRoutersail-research/fp4	$2.6	$13	Source09/06/2026
	Together	Via OpenRoutertogether	$3	$15	Source09/06/2026
	Wafer	Via OpenRouterwafer	$3	Lowest listed price: $12.75	Source09/06/2026

Performance

Shared benchmarks

Only published results with the same benchmark version, task, metric, and comparison cohort are paired. Different reasoning levels, agents, harnesses, or output limits appear directly in the table.

Capability profile from matched benchmarks

Each axis averages directly matched benchmark families. A value of 100 means the stronger value within this pair, not a universal quality score.

Grok 4.6

Kimi K3

Coding: 97.4 / 83.2 (4 benchmark families)

Reasoning: 100 / 98.1 (1 benchmark family)

Mathematics: 58.6 / 100 (1 benchmark family)

Knowledge: 100 / 98.4 (1 benchmark family)

Finance: 95.3 / 100 (5 benchmark families)

Legal: 100 / 86.7 (3 benchmark families)

Agentic tasks: 98.4 / 94.3 (2 benchmark families)

Productivity: 97.9 / 100 (1 benchmark family)

23 matched rows from 23 benchmark families were considered. The radar shows 8 of 11 comparable categories.

Grok 4.6Kimi K3

BioMysteryBenchOverall. Higher is better. No documented setup difference

Grok 4.672.222 %

Kimi K371.481 %

Code MigrationOverall. Higher is better. No documented setup difference

Grok 4.644.572 %

Kimi K316.099 %

CorpFin v2Overall. Higher is better. No documented setup difference

Grok 4.666.161 %

Kimi K371.562 %

EMBOverall. Higher is better. No documented setup difference

Grok 4.662.727 %

Kimi K366.399 %

Finance Agent (v2)Overall. Higher is better. No documented setup difference

Grok 4.653.682 %

Kimi K354.36 %

GPQA DiamondOverall. Higher is better. No documented setup difference

Grok 4.694.697 %

Kimi K392.929 %

Harvey's Legal Agent BenchmarkOverall · Task fully resolved. Higher is better. No documented setup difference

Grok 4.615.833 %

Kimi K310.833 %

Legal Research BenchOverall · All-pass. Higher is better. No documented setup difference

Grok 4.648.077 %

Kimi K344.231 %

LegalBenchOverall. Higher is better. No documented setup difference

Grok 4.686.307 %

Kimi K386.016 %

LiveCodeBench v6Overall. Higher is better. No documented setup difference

Grok 4.688.224 %

Kimi K387.185 %

LMArena Text, Style ControlOverall. Higher is better. Test setup not documented as identical

Grok 4.61,461.115

Kimi K31,489.017

MedCodeOverall. Higher is better. No documented setup difference

Grok 4.644.712 %

Kimi K348.884 %

Key to the marks:better valuebetter, but not conclusivetie

Benchmark scores for Grok 4.6 and Kimi K3
Benchmark and task	Grok 4.6	Kimi K3	Source
BioMysteryBenchOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Harness: Terminus 2	Winner72.222 %	71.481 %	Vals AI Source details Data as of: 09/02/2026
Code MigrationOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner44.572 %	16.099 %	Vals AI Source details Data as of: 09/03/2026
CorpFin v2Overall Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is better	66.161 %	Winner71.562 %	Vals AI Source details Data as of: 08/12/2026
EMBOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	62.727 %	Winner66.399 %	Vals AI Source details Data as of: 09/03/2026
Finance Agent (v2)Overall Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is better	53.682 %	Winner54.36 %	Vals AI Source details Data as of: 09/03/2026
GPQA DiamondOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner94.697 %	92.929 %	Vals AI Source details Data as of: 09/01/2026
Harvey's Legal Agent BenchmarkOverall · Task fully resolved Test details No documented difference in the test setupVersion: 1Metric: task resolution rateScoring: Higher is better	Winner15.833 %	10.833 %	Vals AI Source details Data as of: 09/03/2026
Legal Research BenchOverall · All-pass Test details No documented difference in the test setupVersion: 1Metric: all-pass rateScoring: Higher is better	Winner48.077 %	44.231 %	Vals AI Source details Data as of: 09/03/2026
LegalBenchOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner86.307 %	86.016 %	Vals AI Source details Data as of: 09/01/2026
LiveCodeBench v6Overall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner88.224 %	87.185 %	Vals AI Source details Data as of: 09/01/2026
LMArena Text, Style ControlOverall Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 3,453 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 17,895 anonymous pairwise votes.	1,461.115 95% confidence interval: ±10.144The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,489.017 95% confidence interval: ±5.47The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
MedCodeOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	44.712 %	Winner48.884 %	Vals AI Source details Data as of: 09/03/2026
MedScribeOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	86.534 %	Winner87.958 %	Vals AI Source details Data as of: 09/03/2026
MMLU-ProOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner89.4 %	87.973 %	Vals AI Source details Data as of: 09/01/2026
MortgageTaxOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	64.189 %	Winner66.336 %	Vals AI Source details Data as of: 09/01/2026
ProofBenchOverall Test details No documented difference in the test setupVersion: 1.1Metric: accuracyScoring: Higher is better	51 %	Winner87 %	Vals AI Source details Data as of: 09/01/2026
Public Benefits Bench v1.1Overall Test details No documented difference in the test setupVersion: 1.1Metric: accuracyScoring: Higher is better	66.847 %	Winner68.268 %	Vals AI Source details Data as of: 09/01/2026
SAGEOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	28.9 %	Winner54.26 %	Vals AI Source details Data as of: 09/03/2026
SWE-bench VerifiedOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner95.6 %	93.4 %	Vals AI Source details Data as of: 09/01/2026
TaxEval v2Overall Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is better	71.096 %	Winner75.716 %	Vals AI Source details Data as of: 09/01/2026
Terminal-Bench 2.1Overall Test details No documented difference in the test setupVersion: 2.1Metric: accuracyScoring: Higher is better	78.277 %	Winner80.899 %	Vals AI Source details Data as of: 09/03/2026
Time Horizon Index: KSPOverall Test details No documented difference in the test setupVersion: 1Metric: progressScoring: Higher is better	Winner5.833 %	5.167 %	Vals AI Source details Data as of: 08/20/2026
Vals IndexOverall Test details No documented difference in the test setupVersion: 2Metric: weighted index scoreScoring: Higher is better	Winner59.167 %	57.813 %	Vals AI Source details Data as of: 09/03/2026
Vibe Code Bench v1.1Overall Test details No documented difference in the test setupVersion: 1.1Metric: accuracyScoring: Higher is betterSettings: Harness: OpenHands	76.239 %	Winner84.963 %	Vals AI Source details Data as of: 09/03/2026
LMArena Text, Style ControlBusiness, management and finance Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 705 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 3,520 anonymous pairwise votes.	1,467.318 95% confidence interval: ±22.146The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,489.216 95% confidence interval: ±10.41The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlChinese Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 291 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 1,168 anonymous pairwise votes.	Higher measured value1,538.386 95% confidence interval: ±33.69The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,528.536 95% confidence interval: ±17.657The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlCoding Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 838 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 4,669 anonymous pairwise votes.	1,510.751 95% confidence interval: ±20.826The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,541.739 95% confidence interval: ±9.193The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlCreative writing Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 651 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 3,543 anonymous pairwise votes.	1,458.096 95% confidence interval: ±23.926The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,459.502 95% confidence interval: ±10.685The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlEnglish Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 1,434 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 6,981 anonymous pairwise votes.	1,459.46 95% confidence interval: ±15.407The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,496.027 95% confidence interval: ±7.574The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlEntertainment, sports and media Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 777 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 4,327 anonymous pairwise votes.	1,443.149 95% confidence interval: ±21.821The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,451.659 95% confidence interval: ±9.734The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlExcluding ties Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 2,522 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 13,279 anonymous pairwise votes.	1,465.128 95% confidence interval: ±14.002The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,502.224 95% confidence interval: ±7.169The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlExpert prompts Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 401 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 1,606 anonymous pairwise votes.	1,479.844 95% confidence interval: ±28.841The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,527.236 95% confidence interval: ±15.094The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlHard prompts Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 2,312 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 11,628 anonymous pairwise votes.	1,484.219 95% confidence interval: ±12.5The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,519.227 95% confidence interval: ±6.437The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlHard prompts, English Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 927 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 4,510 anonymous pairwise votes.	1,475.235 95% confidence interval: ±19.394The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,517.706 95% confidence interval: ±9.203The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlInstruction following Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 1,223 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 5,969 anonymous pairwise votes.	1,458.276 95% confidence interval: ±17.043The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,487.97 95% confidence interval: ±8.244The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlLegal and government Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 284 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 1,419 anonymous pairwise votes.	1,475.664 95% confidence interval: ±33.128The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,509.331 95% confidence interval: ±16.311The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlLife, physical and social science Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 512 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 2,821 anonymous pairwise votes.	1,492.359 95% confidence interval: ±25.684The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,518.8 95% confidence interval: ±11.431The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlLonger queries Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 1,609 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 7,952 anonymous pairwise votes.	1,478.218 95% confidence interval: ±15.132The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,502.353 95% confidence interval: ±7.68The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlMath Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 216 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 756 anonymous pairwise votes.	1,426.012 95% confidence interval: ±40.021The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,504.235 95% confidence interval: ±21.422The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlMathematical professions Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 212 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 892 anonymous pairwise votes.	1,456.037 95% confidence interval: ±39.211The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,513.757 95% confidence interval: ±19.948The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlMedicine and healthcare Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 225 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 1,234 anonymous pairwise votes.	1,460.477 95% confidence interval: ±37.871The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,509.063 95% confidence interval: ±17.314The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlMulti-turn Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 569 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 2,862 anonymous pairwise votes.	1,446.12 95% confidence interval: ±25.079The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,499.114 95% confidence interval: ±11.625The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlNon-English Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 2,019 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 10,909 anonymous pairwise votes.	1,456.123 95% confidence interval: ±13.128The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,476.872 95% confidence interval: ±6.548The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlRussian Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 413 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 1,949 anonymous pairwise votes.	1,446.297 95% confidence interval: ±28.9The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,490.113 95% confidence interval: ±13.716The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlSoftware and IT services Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 1,379 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 7,013 anonymous pairwise votes.	1,503.685 95% confidence interval: ±16.15The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,528.32 95% confidence interval: ±7.648The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
LMArena Text, Style ControlWriting, literature and language Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGrok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style control Kimi K3: Snapshot: kimi-k3-max, Reasoning: max, Harness: Arena text, style controlGrok 4.6: Style-controlled Arena rating from 921 anonymous pairwise votes.Kimi K3: Style-controlled Arena rating from 4,706 anonymous pairwise votes.	1,445.223 95% confidence interval: ±19.746The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,472.249 95% confidence interval: ±9.224The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
Key to the marks:better valuebetter, but not conclusivetie

How to read this comparison

Winning one benchmark is not an overall verdict

Your workload, budget, and required context length matter most. A coding benchmark says little about visual understanding or agent performance.

The standard price table uses direct developer rates. The provider comparison labels OpenRouter endpoints, regions, and special tariffs separately.

Every result links to its measurement source and retrieval date. Labels distinguish vendor reports, official benchmark leaderboards, and independent evaluations. Disputed or archived results are excluded.

More matchups

Compare Grok 4.6 and Kimi K3 with other leading models using the same data and benchmark logic.

View the complete LLM comparison