Gemini 4 Argonvs.DeepSeek V4.1 Flash 1017

Which one is better?

Gemini 4 Argon is clearly ahead in the overall index, but costs 13.3 times (input) and 16.7 times (output) as much per token.

IntelligenceOverall index, 0 to 100 · higher is better

Gemini 4 Argon

Rank 2 to 7

DeepSeek V4.1 Flash 1017

Rank 14 to 22

SpeedTokens per second · higher is better

Gemini 4 Argon

fairly slow

178

DeepSeek V4.1 Flash 1017

fast

PriceDollars per 1M tokens · lower is better

Gemini 4 Argon

fairly expensive

$0.26

DeepSeek V4.1 Flash 1017

cheap

Gemini 4 Argon

Subscription: At Google DeepMind from €21.99 a month (Google AI Pro, with limited credits). Through other providers from ≈ €6.36 a month (ElevenLabs Starter, with limited credits). 28 plans in Germany.
Via API: $2 input, $10 output per 1M tokens.
Note: There is a successor: Gemini 4 Carbon (rank 1 to 5).

DeepSeek V4.1 Flash 1017

Subscription: No DeepSeek subscription includes this model. Through other providers from ≈ €6.36 a month (Token Plan Individual Lite, with limited credits). 7 plans in Germany.
Via API: $0.15 input, $0.60 output per 1M tokens according to the provider; the price index has listed it at $0.30 and $1 since Oct 26, 2026.

Overall index with data through Oct 29, 2026 · Prices from Oct 29, 2026 · Speed: measured by Artificial Analysis · Subscriptions as of Oct 24, 2026 · Subscription prices incl. VAT

Where is the difference?

Values from 0 to 100 per task area. We only name a winner when the gap is large enough and there are enough tests.

Gemini 4 Argon ahead: 5DeepSeek V4.1 Flash 1017 ahead: 0level: 1too little data: 4of 10 areas

AreaGemini 4 ArgonDeepSeek V4.1 Flash 10170 to 100, higher is betterResult

Areas in the overall index

CodingGemini 4 Argon ahead9072

ReasoningGemini 4 Argon ahead9045

Mathematicsonly Gemini 4 Argon has a score90not enough data

KnowledgeGemini 4 Argon ahead9733

Agentslevel8075

Generalonly Gemini 4 Argon has a score92not enough data

Multimodalonly DeepSeek V4.1 Flash 1017 has a scorenot enough data55

Specialist fields

FinanceGemini 4 Argon ahead9269

LegalGemini 4 Argon ahead9176

Healthonly Gemini 4 Argon has a score93not enough data

Show 3 more areas that do not count toward the overall index

AreaGemini 4 ArgonDeepSeek V4.1 Flash 10170 to 100, higher is betterResult

Cybersecurityonly Gemini 4 Argon has a score97not enough data

Productivityonly Gemini 4 Argon has a score91not enough data

Educationonly Gemini 4 Argon has a score90not enough data

See for yourself: same task, one attempt

There is no shared test round for this selection. Results from different rounds are not mixed.

All data side by side

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

Metric	Gemini 4 Argon	DeepSeek V4.1 Flash 1017
Intelligence
Overall index0 to 100, higher is better	93	64
Rank	Rank 2 to 7 of 95	Rank 14 to 22 of 95
Price per 1M tokens
Inputwhat you send	$2	$0.15
Outputwhat the model writes	$10	$0.60
Speed
Tokens per second	82	178
Technical
Context windowhow much text the model reads at once	1.05 M tokens≈ 1,570 A4 pages	1.05 M tokens≈ 1,570 A4 pages
Knowledge cutoff	n/a	n/a
Released	Oct 13, 2026	Oct 17, 2026
Availability	Active	Open Source
Open weights	no	yes

Gemini 4 Argon, Standard pricing through November 30, 2026: Per 1 million tokens: $2.00 input, $10.00 output including thinking, and $0.10 cached input. Source
Gemini 4 Argon, Standard pricing from December 1, 2026: Per 1 million tokens: $4.00 input, $20.00 output including thinking, and $0.20 cached input. Source
Gemini 4 Argon, Caching: Implicit caching only; explicit caching is not supported. Source
DeepSeek V4.1 Flash 1017: At peak times (Mon, Tue, Wed, Thu, Fri, 01:00-04:00 UTC; Mon, Tue, Wed, Thu, Fri, 06:00-10:00 UTC) the price is $0.30 input and $1.20 output per 1M tokens. The table shows the off-peak price.
DeepSeek V4.1 Flash 1017, Time-based API pricing: Off peak: $0.15 input, $0.60 output, and $0.003 cached input. Weekdays UTC 01:00-04:00 and 06:00-10:00: $0.30, $1.20, and $0.006. Chinese public holidays are excluded. Source

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

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: Gemini 4 ArgonB: DeepSeek V4.1 Flash 1017

Code MigrationOverall · higher is betterTest setup not documented as identical68.2%44.2%Vals AIAs of 10/18/2026

Code Migration v1.1Overall · higher is betterTest setup not documented as identical70.8%67.5%Vals AIAs of 10/20/2026

EMBOverall · higher is betterTest setup not documented as identical75.2%61.8%Vals AIAs of 10/20/2026

EMB v1.1Overall · higher is betterTest setup not documented as identical74.8%69.2%Vals AIAs of 10/26/2026

Finance Agent (v2)Overall · higher is betterTest setup not documented as identical65.4%57%Vals AIAs of 10/20/2026

Finance Agent v2.1Overall · higher is betterTest setup not documented as identical67.7%62.9%Vals AIAs of 10/26/2026

HLAB-Vals-1.1Overall · Task fully resolved · higher is betterTest setup not documented as identical27.1%14.2%Vals AIAs of 10/26/2026

Legal Research BenchOverall · All-pass · higher is betterTest setup not documented as identical54.8%45.2%Vals AIAs of 10/26/2026

LMArena Code Arena WebDevOverall · higher is betterTest setup not documented as identical1,6801,666LMArenaAs of 10/27/2026

Tax Agent BenchOverall · higher is betterTest setup not documented as identical44.6%44.4%Vals AIAs of 10/26/2026

Show 54 more benchmarks

Terminal-Bench 4.0Overall · higher is betterTest setup not documented as identical57.6%31.8%Vals AIAs of 10/26/2026

Vals Finance IndexOverall · higher is betterTest setup not documented as identical64.9%60.9%Vals AIAs of 10/26/2026

Vals IndexOverall · higher is betterTest setup not documented as identical69.2%61.3%Vals AIAs of 10/27/2026

Vals Legal IndexOverall · higher is betterTest setup not documented as identical76.7%71.9%Vals AIAs of 10/26/2026

Vals Router IndexOverall · higher is better69.2%61.3%Vals AIAs of 10/20/2026

Vibe Code Bench v1.1Overall · higher is betterTest setup not documented as identical91.9%87.7%Vals AIAs of 10/26/2026

Code MigrationCLI · higher is betterTest setup not documented as identical67.6%35.6%Vals AIAs of 10/18/2026

Code Migration v1.1CLI · higher is betterTest setup not documented as identical67.8%59.2%Vals AIAs of 10/20/2026

EMBComps · higher is betterTest setup not documented as identical83%49.8%Vals AIAs of 10/20/2026

EMB v1.1Comps · higher is betterTest setup not documented as identical82.9%70.8%Vals AIAs of 10/26/2026

Finance Agent (v2)Adjustments · higher is betterTest setup not documented as identical60.5%58.4%Vals AIAs of 10/20/2026

Finance Agent v2.1Adjustments · higher is betterTest setup not documented as identical70.8%66.8%Vals AIAs of 10/26/2026

HLAB-Vals-1.1Antitrust Competition · Criteria passed · higher is betterTest setup not documented as identical97.6%93.5%Vals AIAs of 10/26/2026

Legal Research BenchAdministrative / Regulatory · All-pass · higher is betterTest setup not documented as identical68.2%59.1%Vals AIAs of 10/26/2026

LMArena Code Arena WebDevHTML development · higher is betterTest setup not documented as identical1,6731,705LMArenaAs of 10/27/2026

Tax Agent BenchControversy & Precedence Analysis · higher is betterTest setup not documented as identical20%23.3%Vals AIAs of 10/26/2026

Terminal-Bench 4.0Hardware · higher is betterTest setup not documented as identical46.7%6.7%Vals AIAs of 10/26/2026

Vals Finance IndexEMB v1.1 · higher is betterTest setup not documented as identical74.8%69.2%Vals AIAs of 10/26/2026

Vals IndexCode Migration v1.1 · higher is betterTest setup not documented as identical74.4%68.5%Vals AIAs of 10/27/2026

Vals Legal IndexBusiness & Commercial · higher is betterTest setup not documented as identical71.5%68%Vals AIAs of 10/26/2026

Vals Router IndexCode Migration · higher is better74.4%68.5%Vals AIAs of 10/20/2026

Code MigrationCOBOL · higher is betterTest setup not documented as identical70%70%Vals AIAs of 10/18/2026

Code Migration v1.1COBOL · higher is betterTest setup not documented as identical79.7%92.5%Vals AIAs of 10/20/2026

EMBDataroom · higher is betterTest setup not documented as identical85.1%82.8%Vals AIAs of 10/20/2026

EMB v1.1Dataroom · higher is betterTest setup not documented as identical85%86.5%Vals AIAs of 10/26/2026

Finance Agent (v2)All-Pass · higher is betterTest setup not documented as identical55.8%45.2%Vals AIAs of 10/20/2026

Finance Agent v2.1All-Pass · higher is betterTest setup not documented as identical58.9%51.3%Vals AIAs of 10/26/2026

HLAB-Vals-1.1Antitrust Competition · Task fully resolved · higher is betterTest setup not documented as identical60%30%Vals AIAs of 10/26/2026

Legal Research BenchAdministrative / Regulatory · Weighted partial credit · higher is betterTest setup not documented as identical92.9%91.3%Vals AIAs of 10/26/2026

LMArena Code Arena WebDevReact development · higher is betterTest setup not documented as identical1,6941,672LMArenaAs of 10/27/2026

Tax Agent BenchControversy & Precedence Analysis · higher is betterTest setup not documented as identical63.6%73.2%Vals AIAs of 10/26/2026

Terminal-Bench 4.0Media · higher is betterTest setup not documented as identical83.3%25%Vals AIAs of 10/26/2026

Vals Finance IndexFinance Agent v2.1 · higher is betterTest setup not documented as identical67.7%62.9%Vals AIAs of 10/26/2026

Vals IndexEMB v1.1 · higher is betterTest setup not documented as identical74.8%69.2%Vals AIAs of 10/27/2026

Vals Legal IndexCivil Negligence · higher is betterTest setup not documented as identical74%64.6%Vals AIAs of 10/26/2026

Vals Router IndexCoding · higher is better74.6%62.7%Vals AIAs of 10/20/2026

Code MigrationCode Quality · higher is betterTest setup not documented as identical3.574 / 53.992 / 5Vals AIAs of 10/18/2026

Code Migration v1.1Code Quality · higher is betterTest setup not documented as identical3%3.5%Vals AIAs of 10/20/2026

EMBDCF · higher is betterTest setup not documented as identical65.7%53%Vals AIAs of 10/20/2026

EMB v1.1DCF · higher is betterTest setup not documented as identical65.5%59.7%Vals AIAs of 10/26/2026

Finance Agent (v2)Comparables · higher is betterTest setup not documented as identical51%47.5%Vals AIAs of 10/20/2026

Finance Agent v2.1Comparables · higher is betterTest setup not documented as identical61.7%59.8%Vals AIAs of 10/26/2026

HLAB-Vals-1.1Arbitration International Dispute Resolution · Criteria passed · higher is betterTest setup not documented as identical96.1%91.8%Vals AIAs of 10/26/2026

Legal Research BenchBusiness & Commercial · All-pass · higher is betterTest setup not documented as identical46.5%44.2%Vals AIAs of 10/26/2026

LMArena Code Arena WebDevWeb development · higher is betterTest setup not documented as identical1,6801,666LMArenaAs of 10/27/2026

Tax Agent BenchCurrent & Temporal Analysis · higher is betterTest setup not documented as identical66.7%56.4%Vals AIAs of 10/26/2026

Terminal-Bench 4.0ML · higher is betterTest setup not documented as identical69.7%42.4%Vals AIAs of 10/26/2026

Vals Finance IndexTax Agent Bench · higher is betterTest setup not documented as identical44.6%44.4%Vals AIAs of 10/26/2026

Vals IndexFinance Agent v2.1 · higher is betterTest setup not documented as identical67.7%62.9%Vals AIAs of 10/27/2026

Vals Legal IndexCriminal · higher is betterTest setup not documented as identical75.8%62.5%Vals AIAs of 10/26/2026

Vals Router IndexEMB · higher is better74.8%69.2%Vals AIAs of 10/20/2026

EMBLBO · higher is betterTest setup not documented as identical68.5%57.8%Vals AIAs of 10/20/2026

EMB v1.1LBO · higher is betterTest setup not documented as identical67.2%62.5%Vals AIAs of 10/26/2026

Finance Agent (v2)Disclosure Analysis · higher is betterTest setup not documented as identical74.5%66.3%Vals AIAs of 10/20/2026

Compare another matchup

First modelSecond model

More matchups

Compare Gemini 4 Argon and DeepSeek V4.1 Flash 1017 with other leading models.

View the complete LLM comparison

Gemini 4 Argonvs.DeepSeek V4.1 Flash 1017

Gemini 4 Argon

DeepSeek V4.1 Flash 1017

Where is the difference?

See for yourself: same task, one attempt

All data side by side

Shared benchmarks

Compare another matchup

Related LLM comparisons