Head-to-head comparison
GPT-5.6 Luna vs. Mistral Medium 3.5
Compare published benchmark results from matching versions and cohorts, API pricing, and technical specifications. Any documented test setup differences remain visible.
28
shared result series
1
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-5.6 Luna | Mistral Medium 3.5 |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Released | Jun 26, 2026 | Apr 28, 2026 |
| Availability | Active | Active |
| Model type | Proprietary | Open Weights |
| Parameters | Not published | Not published |
| Architecture | Not published | Not published |
| Context window | 1,050,000 tokens | 256,000 tokens |
| Knowledge cutoff | February 16, 2026 | Not published |
| Technical sources | Model, Context, Knowledge cutoff | Model |
| API pricing per 1M tokens | ||
| API input | $0.2 ≤272K / $0.4 >272K | $1.5 |
| API output | $1.2 ≤272K / $1.8 >272K | $7.5 |
| Cache read | $0.02 ≤272K / $0.04 >272K | Not listed |
| Cache write (5 min.) | $0.25 ≤272K / $0.5 >272K | Not listed |
| Cache write (1 hr.) | Not listed | Not listed |
| Price verified | 08/11/2026Source | 07/30/2026Source |
GPT-5.6 LunaMistral Medium 3.5
API inputUSD per 1M tokens, base tier
GPT-5.6 Luna$0.2Lower price
Mistral Medium 3.5$1.5
API outputUSD per 1M tokens, base tier
GPT-5.6 Luna$1.2Lower price
Mistral Medium 3.5$7.5
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.
28 shared result series
GPT-5.6 LunaMistral Medium 3.5
Arena Text, Style ControlOverall. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,450.561 arena pointsHigher measured value
Mistral Medium 3.51,427.008 arena points
Arena Text, Style ControlBusiness, management and finance. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,465.32 arena pointsHigher measured value
Mistral Medium 3.51,430.152 arena points
Arena Text, Style ControlChinese. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,485.14 arena pointsHigher measured value
Mistral Medium 3.51,450.261 arena points
Arena Text, Style ControlCoding. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,498.496 arena pointsHigher measured value
Mistral Medium 3.51,478.733 arena points
Arena Text, Style ControlCreative writing. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,413.67 arena pointsHigher measured value
Mistral Medium 3.51,396.218 arena points
Arena Text, Style ControlEnglish. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,467.903 arena pointsHigher measured value
Mistral Medium 3.51,444.801 arena points
Arena Text, Style ControlEntertainment, sports and media. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,425.084 arena pointsHigher measured value
Mistral Medium 3.51,396.264 arena points
Arena Text, Style ControlExcluding ties. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,453.512 arena pointsHigher measured value
Mistral Medium 3.51,421.102 arena points
Arena Text, Style ControlExpert prompts. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,485.913 arena pointsHigher measured value
Mistral Medium 3.51,437.963 arena points
Arena Text, Style ControlFrench. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,457.035 arena pointsHigher measured value
Mistral Medium 3.51,437.162 arena points
Arena Text, Style ControlGerman. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,499.364 arena pointsHigher measured value
Mistral Medium 3.51,441.705 arena points
Arena Text, Style ControlHard prompts. Higher is better. Test setup not documented as identical
GPT-5.6 Luna1,470.97 arena pointsHigher measured value
Mistral Medium 3.51,445.368 arena points
| Benchmark and task | GPT-5.6 Luna | Mistral Medium 3.5 | Comparability | Source |
|---|---|---|---|---|
| Arena Text, Style ControlOverall | 1,450.561 arena pointsHigher measured value 95% confidence interval: ±5.636The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,427.008 arena points 95% confidence interval: ±6.567The 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:overallTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 16,419 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 11,022 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlBusiness, management and finance | 1,465.32 arena pointsHigher measured value 95% confidence interval: ±11.098The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,430.152 arena points 95% confidence interval: ±13.219The 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_operationsTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 3,171 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 2,143 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlChinese | 1,485.14 arena pointsHigher measured value 95% confidence interval: ±18.537The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,450.261 arena points 95% confidence interval: ±26.734The 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:chineseTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 1,179 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 502 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlCoding | 1,498.496 arena pointsHigher measured value 95% confidence interval: ±9.171The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,478.733 arena points 95% confidence interval: ±11.154The 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:codingTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 4,743 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 3,082 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlCreative writing | 1,413.67 arena pointsHigher measured value 95% confidence interval: ±11.521The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,396.218 arena points 95% confidence interval: ±14.428The 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_writingTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 3,074 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 1,861 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlEnglish | 1,467.903 arena pointsHigher measured value 95% confidence interval: ±7.784The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,444.801 arena points 95% confidence interval: ±8.671The 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:englishTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 6,691 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 5,311 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlEntertainment, sports and media | 1,425.084 arena pointsHigher measured value 95% confidence interval: ±10.093The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,396.264 arena points 95% confidence interval: ±12.523The 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_mediaTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 4,027 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 2,442 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlExcluding ties | 1,453.512 arena pointsHigher measured value 95% confidence interval: ±7.354The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,421.102 arena points 95% confidence interval: ±8.693The 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_tiesTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 12,388 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 8,302 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlExpert prompts | 1,485.913 arena pointsHigher measured value 95% confidence interval: ±14.333The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,437.963 arena points 95% confidence interval: ±17.885The 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:expertTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 1,786 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 1,109 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlFrench | 1,457.035 arena pointsHigher measured value 95% confidence interval: ±25.74The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,437.162 arena points 95% confidence interval: ±30.614The 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:frenchTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 579 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 440 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlGerman | 1,499.364 arena pointsHigher measured value 95% confidence interval: ±36.569The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,441.705 arena points 95% confidence interval: ±44.46The 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:germanTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 266 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 170 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlHard prompts | 1,470.97 arena pointsHigher measured value 95% confidence interval: ±6.62The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,445.368 arena points 95% confidence interval: ±7.871The 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_promptsTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 11,043 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 7,159 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlHard prompts, English | 1,486.256 arena pointsHigher measured value 95% confidence interval: ±9.271The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,457.822 arena points 95% confidence interval: ±10.361The 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_englishTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 4,500 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 3,642 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlInstruction following | 1,442.032 arena pointsHigher measured value 95% confidence interval: ±8.307The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,422.086 arena points 95% confidence interval: ±10.353The 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_followingTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 5,883 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 3,557 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlKorean | 1,386.735 arena points 95% confidence interval: ±30.324The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,392.952 arena pointsHigher measured value 95% confidence interval: ±44.647The 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:koreanTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 363 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 178 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlLegal and government | 1,460.755 arena pointsHigher measured value 95% confidence interval: ±16.612The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,438.388 arena points 95% confidence interval: ±21.256The 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_legal_and_governmentTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 1,372 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 844 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 science | 1,454.714 arena pointsHigher measured value 95% confidence interval: ±11.91The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,433.414 arena points 95% confidence interval: ±14.39The 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_scienceTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 2,634 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 1,823 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlLonger queries | 1,450.194 arena pointsHigher measured value 95% confidence interval: ±7.794The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,430.645 arena points 95% confidence interval: ±9.453The 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_queryTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 7,739 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 4,657 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlMath | 1,484.009 arena pointsHigher measured value 95% confidence interval: ±21.054The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,429.157 arena points 95% confidence interval: ±24.882The 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:mathTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 764 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 539 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlMathematical professions | 1,477.929 arena pointsHigher measured value 95% confidence interval: ±20.986The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,439.68 arena points 95% confidence interval: ±24.051The 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_mathematicalTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 835 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 590 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlMedicine and healthcare | 1,433.741 arena pointsHigher measured value 95% confidence interval: ±17.642The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,423.156 arena points 95% confidence interval: ±21.773The 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_medicine_and_healthcareTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 1,234 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 807 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlMulti-turn | 1,455.655 arena pointsHigher measured value 95% confidence interval: ±11.753The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,431.607 arena points 95% confidence interval: ±14.403The 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_turnTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 2,815 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 1,812 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlNon-English | 1,433.062 arena pointsHigher measured value 95% confidence interval: ±6.88The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,406.489 arena points 95% confidence interval: ±8.386The 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_englishTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 9,725 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 5,711 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlPolish | 1,430.299 arena pointsHigher measured value 95% confidence interval: ±34.831The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,408.19 arena points 95% confidence interval: ±38.902The 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:polishTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 285 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 226 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlRussian | 1,449.175 arena pointsHigher measured value 95% confidence interval: ±14.826The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,416.701 arena points 95% confidence interval: ±17.895The 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:russianTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 1,654 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 1,140 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlSoftware and IT services | 1,490.948 arena pointsHigher measured value 95% confidence interval: ±7.94The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,470.303 arena points 95% confidence interval: ±9.539The 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_servicesTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 6,687 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 4,328 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlSpanish | 1,450.132 arena pointsHigher measured value 95% confidence interval: ±30.269The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,387.211 arena points 95% confidence interval: ±32.775The 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:spanishTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 413 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 373 anonymous pairwise votes. | ArenaType: Official benchmark leaderboardLeaderboard updated: 2026-08-12Import ID: lmarena-eeedf9209373a7f9b324bfb839e7b02fedacd1e5-1611ed64f3f1Retrieved 2026-08-16 |
| Arena Text, Style ControlWriting, literature and language | 1,439.47 arena pointsHigher measured value 95% confidence interval: ±9.762The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result. | 1,405.854 arena points 95% confidence interval: ±11.763The 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_languageTest setup not documented as identicalNo direct winnerGPT-5.6 Luna: Snapshot: gpt-5.6-luna-xhigh, Reasoning: xhigh, Harness: Arena text, style control Mistral Medium 3.5: Snapshot: mistral-medium-3.5, Harness: Arena text, style controlGPT-5.6 Luna: Style-controlled Arena rating from 4,231 anonymous pairwise votes.Mistral Medium 3.5: Style-controlled Arena rating from 2,730 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-5.6 Luna and Mistral Medium 3.5 with other leading models using the same data and benchmark logic.
View the complete LLM comparison