OpenAIDeepSeekLLM matchup

GPT-6 Lunavs.DeepSeek-V4-Flash

Which one is better?

DeepSeek-V4-Flash does not have enough benchmark results yet. GPT-6 Luna is cheaper for input, DeepSeek-V4-Flash for output.

IntelligenceOverall index, 100 points = average

96.3

GPT-6 Luna

Rank 35

n/a

DeepSeek-V4-Flash

not ranked yet

DeepSeek-V4-Flash faster

SpeedTokens per second

101

GPT-6 Luna

123

DeepSeek-V4-Flash

about the same price

Price$ per 1M tokens

$0.20

GPT-6 Luna

$0.18

DeepSeek-V4-Flash

about as popular

Popularity0 to 100

GPT-6 Luna

Rank 30 of 785

DeepSeek-V4-Flash

Rank 17 of 785

Above 272,000 tokens of context, GPT-6 Luna costs more: $0.20 input and $0.75 output per 1M tokens.

Choose GPT-6 Luna for

Agents

Choose DeepSeek-V4-Flash for

No clear advantage

Where is the difference?

GPT-6 Luna is ahead in 1 of 10 areas, DeepSeek-V4-Flash in none.

GPT-6 Luna ahead: 1too little data: 9

Areas in the overall index

AgentsGPT-6 Luna +1699.5 to 83.2

Codingonly GPT-6 Luna has a score

Reasoningonly GPT-6 Luna has a score

Mathematicsonly GPT-6 Luna has a score

Knowledgeonly GPT-6 Luna has a score

Generalonly GPT-6 Luna has a score

Multimodalonly GPT-6 Luna has a score

Specialist fields

Financeonly GPT-6 Luna has a score

Legalonly GPT-6 Luna has a score

Healthonly GPT-6 Luna has a score

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

Show 4 more areas that do not count toward the overall index

Cybersecurityonly GPT-6 Luna has a score

Searchonly GPT-6 Luna has a score

Productivityonly GPT-6 Luna has a score

Educationonly GPT-6 Luna has a score

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

Which model for which task?

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

TaskBetter choicebased on

Writing and debugging codeopenCoding without a score for DeepSeek-V4-Flash

Maths and data analysisopenMathematics without a score for DeepSeek-V4-Flash, Reasoning without a score for DeepSeek-V4-Flash

Facts and researchopenKnowledge without a score for DeepSeek-V4-Flash, Search without a score for DeepSeek-V4-Flash

Agents and workflowsGPT-6 LunaAgents 99.5 to 83.2

Fast chat repliesDeepSeek-V4-Flash101 to 123 tokens per second

Security analysisopenCybersecurity without a score for DeepSeek-V4-Flash

Many requests on a small budgetabout the same$0.20 to $0.18 per 1M tokens

Finance, legal and medical textsopenFinance without a score for DeepSeek-V4-Flash, Legal without a score for DeepSeek-V4-Flash, Health without a score for DeepSeek-V4-Flash

Images, screenshots and PDFsopenMultimodal without a score for DeepSeek-V4-Flash

See for yourself: same task, one attempt

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

Pelican on a bicycle

Prompt

Generate an SVG of a pelican riding a bicycle

Task origin: Simon Willison

GPT-6 Luna

Oct 1, 2026 · 47 s · $0.0022

Test details

Run date: Oct 1, 2026
API provider: openrouter
Requested API model: openai/gpt-6-luna
Reasoning setting: high
Token limit including reasoning: 8,192 tokens
Output tokens: 4,493 tokens
Cost of the attempt: $0.0022
Response time: 46.7 s
API model reported in response: openai/gpt-6-luna
Pinned endpoint: OpenAI | openai/gpt-6-luna-20260922
Temperature: Provider default
Test protocol: community-visual-tests-v1
Technical checks: Passed, not a quality rating

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

DeepSeek-V4-Flash

The response reached the token limit. No complete recording is available.

Oct 5, 2026 · 98 s · $0.0011

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

Test details

Run date: Oct 5, 2026
API provider: openrouter
Requested API model: deepseek/deepseek-v4-flash-0731
Reasoning setting: high
Token limit including reasoning: 8,192 tokens
Output tokens: 8,192 tokens
Cost of the attempt: $0.0011
Response time: 97.8 s
API model reported in response: deepseek/deepseek-v4-flash-0731
Pinned endpoint: StreamLake | deepseek/deepseek-v4-flash-20260731
Temperature: Provider default
Test protocol: community-visual-tests-v1
Technical checks: Failed

Earlier attempts (3)

Oct 2, 2026API request failedReasoning setting: high
Oct 1, 2026API request failedReasoning setting: high
Sep 7, 2026API request failedReasoning setting: high

SVG short test

One SVG task with the same fixed settings for every model, without hidden thinking.

Prompt

Return only one standalone SVG of a pelican riding a bicycle. Use viewBox="0 0 800 600", simple shapes, two clearly visible wheels and a recognizable pelican with a long beak. At most 30 SVG elements and 1800 characters. No Markdown, explanations, comments, scripts, animation, CSS or external resources.

GPT-6 Luna

Unavailable in this round

DeepSeek-V4-Flash

Popularity compared

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

Both are about as popular: GPT-6 Luna 60, DeepSeek-V4-Flash 64 out of 100.

Over time

Weekly values from October 18, 2026 to November 7, 2026

GPT-6 LunaDeepSeek-V4-Flash

All data side by side

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

Metric	GPT-6 Luna	DeepSeek-V4-Flash
Intelligence
Overall indexpoints, 100 = reference group average, higher is better	96.3	n/a
Rank	Rank 35 of 92	not ranked yet
Price per 1M tokens
Inputwhat you send	$0.10above 272,000 tokens: $0.20	$0.14
Outputwhat the model writes	$0.50above 272,000 tokens: $0.75	$0.28
Cost per task (own tests)average of 2 original-protocol tasks both models passed (SVG short test not included)	$0.004	$0.002
Speed
Tokens per second	101	123
Popularity
Popularitysearch, usage and mentions, 0 to 100	60Rank 30 of 785	64Rank 17 of 785
Technical
Context windowhow much text the model reads at once	1.05 M tokens≈ 1,580 A4 pages	1 M tokens≈ 1,500 A4 pages
Knowledge cutoff	May 18, 2026	n/a
Released	Sep 22, 2026	Apr 24, 2026
Availability	Active	Limited access
Open weights	no	yes

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

Is there a better choice?

Successor of DeepSeek-V4-FlashDeepSeek-V4.1-Flash

Overall index 99.7, rank 29$0.15 input and $0.60 output per 1M tokens

GPT-6 Luna vs. DeepSeek-V4.1-Flash

Best model in the overall indexClaude Opus 5.5

Overall index 126.1, rank 1$4 input and $20 output per 1M tokens

GPT-6 Luna vs. Claude Opus 5.5

Shared benchmarks

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

A: GPT-6 LunaB: DeepSeek-V4-Flash

GDPval-AAOverall · higher is better1,444.5 elo1,440.6 eloArtificial AnalysisAs of 11/07/2026

LMArena Code Arena WebDevOverall · higher is betterTest setup not documented as identical1,5791,582LMArenaAs of 10/28/2026

LMArena Text, Style ControlOverall · higher is betterTest setup not documented as identical1,4411,439LMArenaAs of 10/28/2026

LMArena Code Arena WebDevHTML development · higher is betterTest setup not documented as identical1,5951,558LMArenaAs of 10/28/2026

LMArena Text, Style ControlBusiness, management and finance · higher is betterTest setup not documented as identical1,4411,435LMArenaAs of 10/28/2026

LMArena Code Arena WebDevReact development · higher is betterTest setup not documented as identical1,5751,585LMArenaAs of 10/28/2026

LMArena Text, Style ControlChinese · higher is betterTest setup not documented as identical1,4831,480LMArenaAs of 10/28/2026

LMArena Code Arena WebDevWeb development · higher is betterTest setup not documented as identical1,5791,582LMArenaAs of 10/28/2026

LMArena Text, Style ControlCoding · higher is betterTest setup not documented as identical1,5061,480LMArenaAs of 10/28/2026

LMArena Text, Style ControlCreative writing · higher is betterTest setup not documented as identical1,4101,410LMArenaAs of 10/28/2026

Show 24 more benchmarks

LMArena Text, Style ControlEnglish · higher is betterTest setup not documented as identical1,4491,450LMArenaAs of 10/28/2026

LMArena Text, Style ControlEntertainment, sports and media · higher is betterTest setup not documented as identical1,4131,407LMArenaAs of 10/28/2026

LMArena Text, Style ControlExcluding ties · higher is betterTest setup not documented as identical1,4411,436LMArenaAs of 10/28/2026

LMArena Text, Style ControlExpert prompts · higher is betterTest setup not documented as identical1,4891,467LMArenaAs of 10/28/2026

LMArena Text, Style ControlFrench prompts · higher is betterTest setup not documented as identical1,4671,458LMArenaAs of 10/28/2026

LMArena Text, Style ControlGerman prompts · higher is betterTest setup not documented as identical1,4351,434LMArenaAs of 10/28/2026

LMArena Text, Style ControlHard prompts · higher is betterTest setup not documented as identical1,4641,458LMArenaAs of 10/28/2026

LMArena Text, Style ControlHard prompts, English · higher is betterTest setup not documented as identical1,4701,465LMArenaAs of 10/28/2026

LMArena Text, Style ControlInstruction following · higher is betterTest setup not documented as identical1,4381,434LMArenaAs of 10/28/2026

LMArena Text, Style ControlJapanese prompts · higher is betterTest setup not documented as identical1,4211,425LMArenaAs of 10/28/2026

LMArena Text, Style ControlKorean prompts · higher is betterTest setup not documented as identical1,4081,379LMArenaAs of 10/28/2026

LMArena Text, Style ControlLegal and government · higher is betterTest setup not documented as identical1,4481,458LMArenaAs of 10/28/2026

LMArena Text, Style ControlLife, physical and social science · higher is betterTest setup not documented as identical1,4531,459LMArenaAs of 10/28/2026

LMArena Text, Style ControlLonger queries · higher is betterTest setup not documented as identical1,4511,448LMArenaAs of 10/28/2026

LMArena Text, Style ControlMath · higher is betterTest setup not documented as identical1,4471,442LMArenaAs of 10/28/2026

LMArena Text, Style ControlMathematical professions · higher is betterTest setup not documented as identical1,4661,445LMArenaAs of 10/28/2026

LMArena Text, Style ControlMedicine and healthcare · higher is betterTest setup not documented as identical1,4331,460LMArenaAs of 10/28/2026

LMArena Text, Style ControlMulti-turn · higher is betterTest setup not documented as identical1,4411,447LMArenaAs of 10/28/2026

LMArena Text, Style ControlNon-English · higher is betterTest setup not documented as identical1,4291,423LMArenaAs of 10/28/2026

LMArena Text, Style ControlPolish prompts · higher is betterTest setup not documented as identical1,4461,435LMArenaAs of 10/28/2026

LMArena Text, Style ControlRussian prompts · higher is betterTest setup not documented as identical1,4381,436LMArenaAs of 10/28/2026

LMArena Text, Style ControlSoftware and IT services · higher is betterTest setup not documented as identical1,4911,473LMArenaAs of 10/28/2026

LMArena Text, Style ControlSpanish prompts · higher is betterTest setup not documented as identical1,4421,430LMArenaAs of 10/28/2026

LMArena Text, Style ControlWriting, literature and language · higher is betterTest setup not documented as identical1,4241,422LMArenaAs of 10/28/2026

More comparisons

Compare GPT-6 Luna and DeepSeek-V4-Flash with the models people compare them with most often.

GPT-6 Lunavs.DeepSeek-V4-Flash

Choose GPT-6 Luna for

Choose DeepSeek-V4-Flash for

Where is the difference?

Which model for which task?

See for yourself: same task, one attempt

Pelican on a bicycle

SVG short test

GPT-6 Luna

DeepSeek-V4-Flash

Popularity compared

Over time

All data side by side

Is there a better choice?

Shared benchmarks

More comparisons

Compare another matchup

Predecessor and successor

The most popular comparisons