Head-to-head comparison

DeepSeek-V4-Pro vs. DeepSeek-V4-Flash

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

shared result series: 61
distinct benchmarks: 2
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 DeepSeek-V4-Pro and DeepSeek-V4-Flash
Attribute	DeepSeek-V4-Pro	DeepSeek-V4-Flash
Provider	DeepSeek	DeepSeek
Released	Apr 24, 2026	Apr 24, 2026
Availability	Preview	Preview
Model type	Open Weights	Open Weights
Parameters	1.6T, 49B active	284B, 13B active
Architecture	Mixture of Experts	Mixture of Experts
Context window	1,000,000 tokens	1,000,000 tokens
Knowledge cutoff	Not published	Not published
Technical sources	Model, Context	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 DeepSeek-V4-Pro and DeepSeek-V4-Flash
Price type	DeepSeek-V4-Pro	DeepSeek-V4-Flash
API input	$0.66	$0.22
API output	$1.98	$0.66
Cache read	$0.02	$0.01
Cache write (5 min.)	Not listed	Not listed
Cache write (1 hr.)	Not listed	Not listed
Price verified	09/08/2026 Source	09/08/2026 Source

DeepSeek-V4-ProDeepSeek-V4-Flash

API inputUSD per 1M tokens, base tier

DeepSeek-V4-Pro$0.66

DeepSeek-V4-Flash$0.22

API outputUSD per 1M tokens, base tier

DeepSeek-V4-Pro$1.98

DeepSeek-V4-Flash$0.66

Cache readUSD per 1M tokens, base tier

DeepSeek-V4-Pro$0.022

DeepSeek-V4-Flash$0.007

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 DeepSeek-V4-Pro and DeepSeek-V4-Flash
Model	Provider	Purchase route and tariff	Input per 1M tokens	Output per 1M tokens	Source
DeepSeek-V4-Pro	DeepSeek	Direct from developerStandard API	Lowest listed price: $0.66	Lowest listed price: $1.98	Source09/08/2026
	Alibaba	Via OpenRouteralibaba	$1.122	$3.366	Source09/08/2026
	Baidu	Via OpenRouterbaidu/fp8	$1.1194	$3.3581	Source09/08/2026
	BaseTen	Via OpenRouterbaseten/fp4#1	$1.32	$3.96	Source09/08/2026
	BaseTen	Via OpenRouterbaseten/fp4#2	$1.32	$3.96	Source09/08/2026
	Cloudflare	Via OpenRoutercloudflare	$1.32	$3.96	Source09/08/2026
	CoreWeave	Via OpenRoutercoreweave/fp8	$1.31	$3.96	Source09/08/2026
	DeepInfra	Via OpenRouterdeepinfra/fp8	$1.3	$2.6	Source09/08/2026
	DeepSeek	Via OpenRouterdeepseek	Lowest listed price: $0.66	$1.98	Source09/08/2026
	DigitalOcean	Via OpenRouterdigitalocean	$1.32	$3.96	Source09/08/2026
	Fireworks	Via OpenRouterfireworks	$1.32	$3.96	Source09/08/2026
	GMICloud	Via OpenRoutergmicloud/fp8	$1.056	$3.168	Source09/08/2026
	Ionstream	Via OpenRouterionstream/fp4	$1.1	$3.3	Source09/08/2026
	NextBit	Via OpenRouternextbit/fp8	$1.122	$3.366	Source09/08/2026
	Novita	Via OpenRouternovita/fp8	$0.99	$2.97	Source09/08/2026
	Parasail	Via OpenRouterparasail/fp8	$1.32	$3.96	Source09/08/2026
	Phala	Via OpenRouterphala	$1.45	$4.36	Source09/08/2026
	Sail Research	Via OpenRoutersail-research/fp4	$1.32	$3.96	Source09/08/2026
	SiliconFlow	Via OpenRoutersiliconflow/fp8	$1.32	$3.96	Source09/08/2026
	StreamLake	Via OpenRouterstreamlake	$1.0494	$3.1482	Source09/08/2026
	Together	Via OpenRoutertogether	$1.32	$3.96	Source09/08/2026
DeepSeek-V4-Flash	DeepSeek	Direct from developerStandard API	$0.22	$0.66	Source09/08/2026
	Alibaba	Via OpenRouteralibaba	$0.352	$1.056	Source09/08/2026
	AtlasCloud	Via OpenRouteratlas-cloud/fp4	$0.44	$1.32	Source09/08/2026
	Baidu	Via OpenRouterbaidu/fp8	$0.14	$0.28	Source09/08/2026
	BaseTen	Via OpenRouterbaseten/fp8#1	$0.13	$0.26	Source09/08/2026
	BaseTen	Via OpenRouterbaseten/fp8#2	$0.13	$0.26	Source09/08/2026
	Cloudflare	Via OpenRoutercloudflare	$0.44	$1.32	Source09/08/2026
	CoreWeave	Via OpenRoutercoreweave/fp8	$0.13	$0.28	Source09/08/2026
	DeepInfra	Via OpenRouterdeepinfra/fp8	$0.06	$0.18	Source09/08/2026
	DeepSeek	Via OpenRouterdeepseek	$0.22	$0.66	Source09/08/2026
	DigitalOcean	Via OpenRouterdigitalocean	$0.08	$0.252	Source09/08/2026
	Fireworks	Via OpenRouterfireworks	$0.22	$0.66	Source09/08/2026
	GMICloud	Via OpenRoutergmicloud/fp8	$0.286	$0.858	Source09/08/2026
	Inceptron	Via OpenRouterinceptron/fp4	$0.13	$0.28	Source09/08/2026
	Makora	Via OpenRoutermakora	$0.09	$0.195	Source09/08/2026
	Mancer 2	Via OpenRoutermancer/fp8	$0.165	$0.5	Source09/08/2026
	Morph	Via OpenRoutermorph/bf16	$0.1234	$0.3475	Source09/08/2026
	NextBit	Via OpenRouternextbit/fp8	$0.352	$1.056	Source09/08/2026
	Novita	Via OpenRouternovita/fp8	$0.4092	$1.2276	Source09/08/2026
	OpenInference	Via OpenRouteropen-inference/fp8	Lowest listed price: $0.05	Lowest listed price: $0.16	Source09/08/2026
	Parasail	Via OpenRouterparasail/fp8	$0.14	$0.28	Source09/08/2026
	Phala	Via OpenRouterphala	$0.44	$1.32	Source09/08/2026
	Reka	Via OpenRouterreka/fp4	$0.11	$0.66	Source09/08/2026
	Relace	Via OpenRouterrelace/fp4	$0.065	$0.18	Source09/08/2026
	Sail Research	Via OpenRoutersail-research/fp4	$0.065	$0.18	Source09/08/2026
	SiliconFlow	Via OpenRoutersiliconflow/fp8	$0.22	$0.66	Source09/08/2026
	StreamLake	Via OpenRouterstreamlake/fp8	$0.132	$0.396	Source09/08/2026
	Together	Via OpenRoutertogether	$0.14	$0.28	Source09/08/2026
	Venice	Via OpenRoutervenice	$0.175	$0.35	Source09/08/2026
	Wafer	Via OpenRouterwafer/fast	$0.1	$0.25	Source09/08/2026

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.

DeepSeek-V4-Pro

DeepSeek-V4-Flash

Coding: 100 / 99.1 (1 benchmark family)

Reasoning: 100 / 98.8 (1 benchmark family)

Mathematics: 100 / 98.2 (1 benchmark family)

Knowledge: 100 / 98.6 (1 benchmark family)

Finance: 100 / 98.7 (1 benchmark family)

Legal: 100 / 98.7 (1 benchmark family)

Cybersecurity: 98.1 / 96 (1 benchmark family)

Health: 100 / 98.7 (1 benchmark family)

61 matched rows from 2 benchmark families were considered. The radar shows 8 of 9 comparable categories.

DeepSeek-V4-ProDeepSeek-V4-Flash

CyberBenchOverall. Higher is better. No documented setup difference

DeepSeek-V4-Pro68.644 %

DeepSeek-V4-Flash66.716 %

LMArena Text, Style ControlOverall. Higher is better. No documented setup difference

DeepSeek-V4-Pro1,457.748

DeepSeek-V4-Flash1,435.816

CyberBenchPatch. Higher is better. No documented setup difference

DeepSeek-V4-Pro79.661 %

DeepSeek-V4-Flash72.414 %

LMArena Text, Style ControlOverall. Higher is better. No documented setup difference

DeepSeek-V4-Pro1,455.047

DeepSeek-V4-Flash1,438.462

CyberBenchPoC. Higher is better. No documented setup difference

DeepSeek-V4-Pro57.627 %

DeepSeek-V4-Flash61.017 %

LMArena Text, Style ControlBusiness, management and finance. Higher is better. No documented setup difference

DeepSeek-V4-Pro1,461.034

DeepSeek-V4-Flash1,437.854

LMArena Text, Style ControlBusiness, management and finance. Higher is better. No documented setup difference

DeepSeek-V4-Pro1,448.251

DeepSeek-V4-Flash1,434.968

LMArena Text, Style ControlChinese. Higher is better. No documented setup difference

DeepSeek-V4-Pro1,493.385

DeepSeek-V4-Flash1,472.677

LMArena Text, Style ControlChinese. Higher is better. No documented setup difference

DeepSeek-V4-Pro1,498.025

DeepSeek-V4-Flash1,479.62

LMArena Text, Style ControlCoding. Higher is better. No documented setup difference

DeepSeek-V4-Pro1,501.769

DeepSeek-V4-Flash1,483.101

LMArena Text, Style ControlCoding. Higher is better. No documented setup difference

DeepSeek-V4-Pro1,488.503

DeepSeek-V4-Flash1,479.449

LMArena Text, Style ControlCreative writing. Higher is better. No documented setup difference

DeepSeek-V4-Pro1,444.463

DeepSeek-V4-Flash1,408.18

Key to the marks:better valuebetter, but not conclusivetie

Our model tests

Community tests you can inspect

Compare how the models respond to the same prompt. Each test shows the first attempt, with no subsequent fixes to the generated code. These results do not contribute to an overall score.

Pelican on a bicycle

A community classic for free-form SVG drawing.

DeepSeek-V4-Pro

API access was unavailable for this run. The model was not evaluated.

First attempt

Run date: 7 Sept 2026
API provider: openrouter
Requested API model: deepseek/deepseek-v4-pro-0813
Pinned endpoint: DeepSeek | deepseek/deepseek-v4-pro-20260813
Reasoning setting: high
Temperature: Provider default
Token limit including reasoning: 8,192 tokens
Test protocol: community-visual-tests-v1
Technical checks: Not run

Not yet visually reviewed. A successful recording does not confirm correct physics or full compliance with the prompt.

DeepSeek-V4-Flash

API access was unavailable for this run. The model was not evaluated.

First attempt

Run date: 7 Sept 2026
API provider: openrouter
Requested API model: deepseek/deepseek-v4-flash-0731
Pinned endpoint: DeepSeek | deepseek/deepseek-v4-flash-20260731
Reasoning setting: high
Temperature: Provider default
Token limit including reasoning: 8,192 tokens
Test protocol: community-visual-tests-v1
Technical checks: Not run

Not yet visually reviewed. A successful recording does not confirm correct physics or full compliance with the prompt.

Prompt and test conditions

Generate an SVG of a pelican riding a bicycle

Token limit including reasoning: 8,192 tokens

Task origin (Simon Willison)

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 DeepSeek-V4-Pro and DeepSeek-V4-Flash with other leading models using the same data and benchmark logic.

View the complete LLM comparison

Benchmark scores for DeepSeek-V4-Pro and DeepSeek-V4-Flash
Benchmark and task	DeepSeek-V4-Pro	DeepSeek-V4-Flash	Source
CyberBenchOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner68.644 %	66.716 %	Vals AI Source details Data as of: 09/01/2026
LMArena Text, Style ControlOverall Test details No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 54,243 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 48,963 anonymous pairwise votes.	Winner1,457.748 95% confidence interval: ±4.009The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,435.816 95% confidence interval: ±4.15The 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
CyberBenchPatch Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner79.661 %	72.414 %	Vals AI Source details Data as of: 08/03/2026
LMArena Text, Style ControlOverall Test details No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 51,599 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 48,705 anonymous pairwise votes.	Winner1,455.047 95% confidence interval: ±4.08The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,438.462 95% confidence interval: ±4.21The 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
CyberBenchPoC Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	57.627 %	Winner61.017 %	Vals AI Source details Data as of: 08/03/2026
LMArena Text, Style ControlBusiness, management and finance Test details No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 10,742 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 9,911 anonymous pairwise votes.	Winner1,461.034 95% confidence interval: ±6.908The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,437.854 95% confidence interval: ±7.128The 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 ControlBusiness, management and finance Test details No documented difference in the test setupNo clear leadVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 10,411 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 10,060 anonymous pairwise votes.	No clear lead1,448.251 95% confidence interval: ±6.986The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,434.968 95% confidence interval: ±7.142The 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 No documented difference in the test setupNo clear leadVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 2,871 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 2,530 anonymous pairwise votes.	No clear lead1,493.385 95% confidence interval: ±11.839The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,472.677 95% confidence interval: ±12.483The 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 No documented difference in the test setupNo clear leadVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 2,757 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 2,509 anonymous pairwise votes.	No clear lead1,498.025 95% confidence interval: ±12.143The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,479.62 95% confidence interval: ±12.459The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 16,165 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 14,346 anonymous pairwise votes.	Winner1,501.769 95% confidence interval: ±6.076The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,483.101 95% confidence interval: ±6.376The 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 No documented difference in the test setupNo clear leadVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 15,113 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 14,310 anonymous pairwise votes.	No clear lead1,488.503 95% confidence interval: ±6.226The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,479.449 95% confidence interval: ±6.409The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 9,051 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 7,972 anonymous pairwise votes.	Winner1,444.463 95% confidence interval: ±7.495The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,408.18 95% confidence interval: ±7.747The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 8,660 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 8,029 anonymous pairwise votes.	Winner1,440.457 95% confidence interval: ±7.584The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,408.903 95% confidence interval: ±7.781The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 23,924 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 21,931 anonymous pairwise votes.	Winner1,467.254 95% confidence interval: ±5.227The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,445.287 95% confidence interval: ±5.383The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 22,700 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 21,747 anonymous pairwise votes.	Winner1,465.768 95% confidence interval: ±5.28The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,449.894 95% confidence interval: ±5.421The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 11,880 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 10,423 anonymous pairwise votes.	Winner1,433.955 95% confidence interval: ±6.804The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,405.349 95% confidence interval: ±7.122The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 11,440 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 10,655 anonymous pairwise votes.	Winner1,430.064 95% confidence interval: ±6.877The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,406.152 95% confidence interval: ±7.06The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 41,089 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 36,983 anonymous pairwise votes.	Winner1,461.499 95% confidence interval: ±5.248The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,431.271 95% confidence interval: ±5.416The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 39,187 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 37,039 anonymous pairwise votes.	Winner1,457.114 95% confidence interval: ±5.263The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,435.7 95% confidence interval: ±5.462The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 5,439 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 4,904 anonymous pairwise votes.	Winner1,481.602 95% confidence interval: ±8.815The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,458.425 95% confidence interval: ±9.147The 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 No documented difference in the test setupNo clear leadVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 5,063 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 4,907 anonymous pairwise votes.	No clear lead1,475.253 95% confidence interval: ±8.974The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,467.744 95% confidence interval: ±9.211The 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 ControlFrench Test details No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 2,116 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 1,909 anonymous pairwise votes.	Winner1,484.137 95% confidence interval: ±15.107The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,444.246 95% confidence interval: ±15.542The 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 ControlFrench Test details No documented difference in the test setupNo clear leadVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 1,986 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 1,872 anonymous pairwise votes.	No clear lead1,470.378 95% confidence interval: ±15.212The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,457.512 95% confidence interval: ±15.76The 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 ControlGerman Test details No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 968 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 834 anonymous pairwise votes.	Winner1,467.776 95% confidence interval: ±19.953The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,423.985 95% confidence interval: ±21.432The 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 ControlGerman Test details No documented difference in the test setupNo clear leadVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 892 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 881 anonymous pairwise votes.	No clear lead1,458.258 95% confidence interval: ±20.588The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,434.001 95% confidence interval: ±20.992The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 36,098 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 32,576 anonymous pairwise votes.	Winner1,480.273 95% confidence interval: ±4.693The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,459.19 95% confidence interval: ±4.879The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 34,551 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 32,479 anonymous pairwise votes.	Winner1,474.049 95% confidence interval: ±4.754The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,458.492 95% confidence interval: ±4.934The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 16,819 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 15,327 anonymous pairwise votes.	Winner1,484.473 95% confidence interval: ±6.001The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,464.306 95% confidence interval: ±6.191The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 15,908 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 15,135 anonymous pairwise votes.	Winner1,482.359 95% confidence interval: ±6.095The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,465.194 95% confidence interval: ±6.261The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 18,672 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 16,772 anonymous pairwise votes.	Winner1,453.345 95% confidence interval: ±5.73The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,428.139 95% confidence interval: ±5.955The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 17,742 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 16,813 anonymous pairwise votes.	Winner1,449.073 95% confidence interval: ±5.821The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,433.908 95% confidence interval: ±6.024The 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 ControlJapanese Test details No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 696 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 633 anonymous pairwise votes.	Winner1,462.265 95% confidence interval: ±24.329The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,397.389 95% confidence interval: ±25.351The 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 ControlJapanese Test details No documented difference in the test setupNo clear leadVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 649 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 600 anonymous pairwise votes.	No clear lead1,430.456 95% confidence interval: ±24.768The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,422.217 95% confidence interval: ±25.99The 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 ControlKorean Test details No documented difference in the test setupNo clear leadVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 1,069 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 867 anonymous pairwise votes.	No clear lead1,418.074 95% confidence interval: ±19.478The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,394.225 95% confidence interval: ±21.434The 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 ControlKorean Test details No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 960 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 909 anonymous pairwise votes.	Winner1,433.057 95% confidence interval: ±20.291The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,374.943 95% confidence interval: ±21.551The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 4,235 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 3,892 anonymous pairwise votes.	Winner1,473.424 95% confidence interval: ±9.866The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,449.174 95% confidence interval: ±10.185The 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 No documented difference in the test setupNo clear leadVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 4,061 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 3,872 anonymous pairwise votes.	No clear lead1,471.913 95% confidence interval: ±10.024The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,459.091 95% confidence interval: ±10.242The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 8,607 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 7,776 anonymous pairwise votes.	Winner1,480.819 95% confidence interval: ±7.337The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,458.608 95% confidence interval: ±7.61The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro-high-preview, Reasoning: high, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash-high-preview, Reasoning: high, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 8,161 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 7,968 anonymous pairwise votes.	Winner1,478.979 95% confidence interval: ±7.429The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,460.882 95% confidence interval: ±7.588The 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 No documented difference in the test setupVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterDeepSeek-V4-Pro: Snapshot: deepseek-v4-pro, Harness: Arena text, style control DeepSeek-V4-Flash: Snapshot: deepseek-v4-flash, Harness: Arena text, style controlDeepSeek-V4-Pro: Style-controlled Arena rating from 23,973 anonymous pairwise votes.DeepSeek-V4-Flash: Style-controlled Arena rating from 21,561 anonymous pairwise votes.	Winner1,472.211 95% confidence interval: ±5.5The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	1,448.456 95% confidence interval: ±5.733The 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