Head-to-head comparison
GPT-5.4 mini vs. Grok 4.6
Compare published benchmark results from matching versions and cohorts, API pricing, and technical specifications. Any documented test setup differences remain visible.
35
shared result series
17
distinct benchmarks
2026-08-16
latest retrieval
Change model selection
Specifications and cost
General specifications and standard API pricing
Undisclosed parameter counts and knowledge cutoffs stay marked as unavailable. The table shows direct standard token pricing, including documented context tiers. Batch, fast, priority, regional surcharges, tool calls, and cloud platform pricing are excluded.
| Attribute | GPT-5.4 mini | Grok 4.6 |
|---|---|---|
| Provider | OpenAI | xAI |
| Released | Mar 2026 | Aug 12, 2026 |
| Availability | Active | Active |
| Model type | Proprietary | Proprietary |
| Parameters | Not published | Not published |
| Architecture | Not published | Not published |
| Context window | 400,000 tokens | 500,000 tokens |
| Knowledge cutoff | August 31, 2025 | February 1, 2026 |
| Technical sources | Model | Model |
| API pricing per 1M tokens | ||
| API input | $0.75 | $2 ≤200K / $4 >200K |
| API output | $4.5 | $6 ≤200K / $12 >200K |
| Cache read | $0.07 | $0.5 ≤200K / $1 >200K |
| Cache write (5 min.) | Not listed | Not listed |
| Cache write (1 hr.) | Not listed | Not listed |
| Price verified | 07/30/2026Source | 08/16/2026Source |
GPT-5.4 miniGrok 4.6
API inputUSD per 1M tokens, base tier
GPT-5.4 mini$0.75Lower price
Grok 4.6$2
API outputUSD per 1M tokens, base tier
GPT-5.4 mini$4.5Lower price
Grok 4.6$6
Cache readUSD per 1M tokens, base tier
GPT-5.4 mini$0.075Lower price
Grok 4.6$0.5
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.
35 shared result series
GPT-5.4 miniGrok 4.6
Arena Text, Style ControlOverall. Higher is better. No documented setup difference
GPT-5.4 mini1,448.205 arena points
Grok 4.61,464.323 arena pointsWinner
Code MigrationOverall. Higher is better. No documented setup difference
GPT-5.4 mini12.936 %
Grok 4.644.572 %Winner
CorpFin v2Overall. Higher is better. No documented setup difference
GPT-5.4 mini60.917 %
Grok 4.666.161 %Winner
EMBOverall. Higher is better. No documented setup difference
GPT-5.4 mini45.426 %
Grok 4.662.727 %Winner
Finance Agent (v2)Overall. Higher is better. No documented setup difference
GPT-5.4 mini45.36 %
Grok 4.653.682 %Winner
GPQA DiamondOverall. Higher is better. No documented setup difference
GPT-5.4 mini83.08 %
Grok 4.694.697 %Winner
Harvey's Legal Agent BenchmarkOverall · Task fully resolved. Higher is better. No documented setup difference
GPT-5.4 mini0 %
Grok 4.615.833 %Winner
Legal Research BenchOverall · All-pass. Higher is better. No documented setup difference
GPT-5.4 mini12.5 %
Grok 4.648.077 %Winner
LiveCodeBench v6Overall. Higher is better. No documented setup difference
GPT-5.4 mini81.465 %
Grok 4.688.224 %Winner
MMLU-ProOverall. Higher is better. No documented setup difference
GPT-5.4 mini84.554 %
Grok 4.689.4 %Winner
MortgageTaxOverall. Higher is better. No documented setup difference
GPT-5.4 mini63.514 %
Grok 4.664.189 %Winner
SAGEOverall. Higher is better. No documented setup difference
GPT-5.4 mini50.813 %Winner
Grok 4.628.9 %
| Benchmark and task | GPT-5.4 mini | Grok 4.6 | Comparability | Source |
|---|---|---|---|---|
| Arena Text, Style ControlOverallReasoning: high | 1,448.205 arena points 95% confidence interval: ±3.859The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,464.323 arena pointsWinner 95% confidence interval: ±11.88The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 59,631 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 2,448 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Code MigrationOverall | 12.936 % | 44.572 %Winner | 1accuracyHigher is betterCohort: vals-code-migration:1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-13Import ID: vals-code-migration-2026-08-16-107652dc6bfcRetrieved 2026-08-16 |
| CorpFin v2Overall | 60.917 % | 66.161 %Winner | 2accuracyHigher is betterCohort: vals-corp-fin-v2:2:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-12Import ID: vals-corp_fin_v2-2026-08-16-45bbf84f57abRetrieved 2026-08-16 |
| EMBOverall | 45.426 % | 62.727 %Winner | 1accuracyHigher is betterCohort: vals-emb:1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-13Import ID: vals-emb-2026-08-16-68d5d1d47a97Retrieved 2026-08-16 |
| Finance Agent (v2)Overall | 45.36 % | 53.682 %Winner | 2accuracyHigher is betterCohort: vals-finance-agent-v2:2:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-14Import ID: vals-fabv2-2026-08-16-8853da43085dRetrieved 2026-08-16 |
| GPQA DiamondOverall | 83.08 % | 94.697 %Winner | 1accuracyHigher is betterCohort: vals-gpqa:1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-12Import ID: vals-gpqa-2026-08-16-7f8363eaa2dbRetrieved 2026-08-16 |
| Harvey's Legal Agent BenchmarkOverall · Task fully resolved | 0 % | 15.833 %Winner | 1task resolution rateHigher is betterCohort: vals-legal-agent-benchmark:1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-14Import ID: vals-hlab-2026-08-16-bae90566df2fRetrieved 2026-08-16 |
| Legal Research BenchOverall · All-pass | 12.5 % | 48.077 %Winner | 1all-pass rateHigher is betterCohort: vals-legal-research:1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-14Import ID: vals-legal_research-2026-08-16-af296c00ecaaRetrieved 2026-08-16 |
| LiveCodeBench v6Overall | 81.465 % | 88.224 %Winner | 1accuracyHigher is betterCohort: vals-lcb:1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-13Import ID: vals-lcb-2026-08-16-17b9c147080dRetrieved 2026-08-16 |
| MMLU-ProOverall | 84.554 % | 89.4 %Winner | 1accuracyHigher is betterCohort: vals-mmlu-pro:1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-12Import ID: vals-mmlu_pro-2026-08-16-e14d5bb3a34cRetrieved 2026-08-16 |
| MortgageTaxOverall | 63.514 % | 64.189 %Winner | 1accuracyHigher is betterCohort: vals-mortgage-tax:1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-12Import ID: vals-mortgage_tax-2026-08-16-e78c985dbb1cRetrieved 2026-08-16 |
| SAGEOverall | 50.813 %Winner | 28.9 % | 1accuracyHigher is betterCohort: vals-sage:1:overallNo documented setup differenceWinner: GPT-5.4 miniGPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-12Import ID: vals-sage-2026-08-16-a416dd25affbRetrieved 2026-08-16 |
| SWE-bench VerifiedOverall | 73 % | 95.6 %Winner | 1accuracyHigher is betterCohort: vals-swebench:1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-12Import ID: vals-swebench-2026-08-16-c133071b8167Retrieved 2026-08-16 |
| TaxEval v2Overall | 71.218 %Winner | 71.096 % | 2accuracyHigher is betterCohort: vals-tax-eval-v2:2:overallNo documented setup differenceWinner: GPT-5.4 miniGPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-12Import ID: vals-tax_eval_v2-2026-08-16-5bbd495dac53Retrieved 2026-08-16 |
| Terminal-Bench 2.1Overall | 54.682 % | 78.277 %Winner | 2.1accuracyHigher is betterCohort: vals-terminal-bench-2-1:2.1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-13Import ID: vals-terminal-bench-2-1-2026-08-16-b3f38c18bea4Retrieved 2026-08-16 |
| Vals IndexOverall | 39.632 % | 59.167 %Winner | 2weighted index scoreHigher is betterCohort: vals-vals-index:2:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17 Grok 4.6: No further settings disclosed | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-14Import ID: vals-vals_index-2026-08-16-ab641d89d6edRetrieved 2026-08-16 |
| Vibe Code Bench v1.1Overall | 47.969 % | 76.239 %Winner | 1.1accuracyHigher is betterCohort: vals-vibe-code-bench:1.1:overallNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-2026-03-17, Harness: OpenHands Grok 4.6: Harness: OpenHands | Vals AIType: Independent evaluationLeaderboard updated: 2026-08-13Import ID: vals-vibe-code-2026-08-16-a73c9770bcc4Retrieved 2026-08-16 |
| Arena Text, Style ControlBusiness, management and financeReasoning: high | 1,460.132 arena pointsNo clear lead 95% confidence interval: ±6.68The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,455.102 arena points 95% confidence interval: ±27.332The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:industry_business_and_management_and_financial_operationsNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 12,114 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 442 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlChineseReasoning: high | 1,478.62 arena points 95% confidence interval: ±11.489The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,535.167 arena pointsWinner 95% confidence interval: ±39.462The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:chineseNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 3,250 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 215 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlCodingReasoning: high | 1,496.777 arena points 95% confidence interval: ±6.045The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,511.77 arena pointsNo clear lead 95% confidence interval: ±26.72The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:codingNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 16,275 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 485 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlCreative writingReasoning: high | 1,404.172 arena points 95% confidence interval: ±7.325The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,472.894 arena pointsWinner 95% confidence interval: ±31.578The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:creative_writingNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 9,849 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 379 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlEnglishReasoning: high | 1,451.541 arena points 95% confidence interval: ±5.021The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,459.266 arena pointsNo clear lead 95% confidence interval: ±18.461The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:englishNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 27,882 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 998 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlEntertainment, sports and mediaReasoning: high | 1,410.773 arena points 95% confidence interval: ±6.698The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,455.476 arena pointsWinner 95% confidence interval: ±28.81The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:industry_entertainment_and_sports_and_mediaNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 12,743 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 448 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlExcluding tiesReasoning: high | 1,450.104 arena points 95% confidence interval: ±5.03The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,469.633 arena pointsNo clear lead 95% confidence interval: ±16.492The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:exclude_tiesNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 45,711 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 1,782 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlExpert promptsReasoning: high | 1,480.978 arena pointsNo clear lead 95% confidence interval: ±8.669The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,479.969 arena points 95% confidence interval: ±34.191The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:expertNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 5,821 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 272 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlHard promptsReasoning: high | 1,470.231 arena points 95% confidence interval: ±4.617The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,488.551 arena pointsNo clear lead 95% confidence interval: ±15.88The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:hard_promptsNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 39,051 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 1,397 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlHard prompts, EnglishReasoning: high | 1,467.272 arena points 95% confidence interval: ±5.757The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,472.181 arena pointsNo clear lead 95% confidence interval: ±25.031The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:hard_prompts_englishNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 19,068 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 543 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlInstruction followingReasoning: high | 1,434.672 arena points 95% confidence interval: ±5.587The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,460.755 arena pointsNo clear lead 95% confidence interval: ±21.635The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:instruction_followingNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 20,246 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 732 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlLife, physical and social scienceReasoning: high | 1,466.152 arena points 95% confidence interval: ±7.094The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,497.908 arena pointsNo clear lead 95% confidence interval: ±31.96The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:industry_life_and_physical_and_social_scienceNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 9,830 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 335 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlLonger queriesReasoning: high | 1,448.515 arena points 95% confidence interval: ±5.438The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,485.451 arena pointsWinner 95% confidence interval: ±18.11The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:longer_queryNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 25,537 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 1,103 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlMulti-turnReasoning: high | 1,469.282 arena pointsNo clear lead 95% confidence interval: ±6.859The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,444.26 arena points 95% confidence interval: ±31.774The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:multi_turnNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 11,195 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 352 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlNon-EnglishReasoning: high | 1,439.067 arena points 95% confidence interval: ±4.821The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,461.99 arena pointsWinner 95% confidence interval: ±15.246The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:non_englishNo documented setup differenceWinner: Grok 4.6GPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 31,748 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 1,450 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlRussianReasoning: high | 1,454.608 arena pointsNo clear lead 95% confidence interval: ±8.482The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,446.737 arena points 95% confidence interval: ±34.404The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:russianNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 6,184 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 275 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlSoftware and IT servicesReasoning: high | 1,486.361 arena points 95% confidence interval: ±5.328The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,511.885 arena pointsNo clear lead 95% confidence interval: ±20.78The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:industry_software_and_it_servicesNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 23,596 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 805 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlWriting, literature and languageReasoning: high | 1,425.266 arena points 95% confidence interval: ±6.234The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,455.393 arena pointsNo clear lead 95% confidence interval: ±25.717The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | style-control-v1style-controlled Bradley-Terry ratingHigher is betterCohort: lmarena-text-style-control:2026-08-12:industry_writing_and_literature_and_languageNo documented setup differenceNo clear leadGPT-5.4 mini: Snapshot: gpt-5.4-mini-high, Reasoning: high, Harness: Arena text, style control Grok 4.6: Snapshot: grok-4.6-high, Reasoning: high, Harness: Arena text, style controlGPT-5.4 mini: Style-controlled Arena rating from 14,591 anonymous pairwise votes.Grok 4.6: Style-controlled Arena rating from 543 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
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.
Prices use official API rates per 1M tokens. Web subscriptions and cloud platform surcharges can differ.
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
Related LLM comparisons
Compare GPT-5.4 mini and Grok 4.6 with other leading models using the same data and benchmark logic.
View the complete LLM comparison