Enter opens the highlighted result. Esc closes.

DeepSeek V4 Flash 0423 vs Mercury 2

DeepSeek V4 Flash 0423 leads Mercury 2 on the Artificial Analysis Intelligence Index, 24.4 to 10.8, and DeepSeek V4 Flash 0423 costs 2.7x less per output token ($0.28 vs $0.75 per million). Updated October 22, 2026; 35 benchmarks have a result for both.

Replace DeepSeek V4 Flash 0423 Replace Mercury 2
DeepSeek V4 Flash 0423DeepSeek
Coding
No data
Overall
No data
Output $/M
$0.28
34
Mercury 2Inception
Coding
No data
Overall
No data
Output $/M
$0.75
1
vs boards won of 35 shared

DeepSeek V4 Flash 0423 or Mercury 2: which to pick

Neither model has a coding score yet, so the two cannot be ranked on coding.

Pick DeepSeek V4 Flash 0423 for output tokens 2.7x cheaper ($0.28 vs $0.75 per million) and a larger context window (1.05M against 128K tokens).

Model ADeepSeek V4 Flash 0423DeepSeekNot benchmarked1 independent benchmarkNo cost per solved taskModel BMercury 2InceptionNot benchmarked1 independent benchmarkNo cost per solved task
Shared results
35
DeepSeek V4 Flash 0423 only
15
Mercury 2 only
7
Like-for-like categories
7 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
DeepSeek V4 Flash 0423 not benchmarked, Mercury 2 not benchmarked.
Widest category
Coding, 29.3 points apart on average over 8 shared benchmarks.
Widest single gap
LMArena WebDev: React: DeepSeek V4 Flash 0423 1585 against Mercury 2 1160, 70.9 points apart on a 0 to 100 scale.
Output price
DeepSeek V4 Flash 0423 $0.28 against Mercury 2 $0.75 per million output tokens, 2.7x apart.
Context window
DeepSeek V4 Flash 0423 takes 1.05M tokens, Mercury 2 128K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryDeepSeek V4 Flash 0423BarsMercury 2Ties
Coding71
Intelligence10
Agentic work10
Reasoning30
Knowledge30
Web design80
General110
All shared boards341
CodingLike-for-likeSix of these boards feed the Vibe Coding Index71
Coding benchmarks for DeepSeek V4 Flash 0423 and Mercury 2
BenchmarkDeepSeek V4 Flash 0423Mercury 2Leader
LMArena WebDev15821171DeepSeek V4 Flash 0423
AA Coding Index52.031.1DeepSeek V4 Flash 0423
SciCodeNot measured37.7%Only Mercury 2 measured
Terminal-Bench Hard (Artificial Analysis run)Not measured26.5%Only Mercury 2 measured
WeirdML45.6%43.2%DeepSeek V4 Flash 0423
ALE-Bench678.2785.6Mercury 2
LMArena Text: coding14571390DeepSeek V4 Flash 0423
LMArena Text: software and IT14531385DeepSeek V4 Flash 0423
LMArena WebDev: React15851160DeepSeek V4 Flash 0423
LMArena WebDev: HTML15581199DeepSeek V4 Flash 0423
Neither model has a result on the other 33 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
Intelligence benchmarks for DeepSeek V4 Flash 0423 and Mercury 2
BenchmarkDeepSeek V4 Flash 0423Mercury 2Leader
AA Intelligence Index24.410.8DeepSeek V4 Flash 0423
Epoch Capabilities Index146.1Not measuredOnly DeepSeek V4 Flash 0423 measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
Agentic work benchmarks for DeepSeek V4 Flash 0423 and Mercury 2
BenchmarkDeepSeek V4 Flash 0423Mercury 2Leader
AA Agentic Index26.31.6DeepSeek V4 Flash 0423
τ²-Bench TelecomNot measured70.8%Only Mercury 2 measured
Neither model has a result on the other 40 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index30
Reasoning benchmarks for DeepSeek V4 Flash 0423 and Mercury 2
BenchmarkDeepSeek V4 Flash 0423Mercury 2Leader
GPQA Diamond (Artificial Analysis run)Not measured77.0%Only Mercury 2 measured
Humanity's Last Exam (Artificial Analysis run)Not measured17.1%Only Mercury 2 measured
SimpleBench46.3%Not measuredOnly DeepSeek V4 Flash 0423 measured
LMArena Text: hard prompts14441362DeepSeek V4 Flash 0423
LMArena Text: expert14411358DeepSeek V4 Flash 0423
LMArena Text: hard prompts in English14491374DeepSeek V4 Flash 0423
Kagi LLM Benchmark52.2%Not measuredOnly DeepSeek V4 Flash 0423 measured
Neither model has a result on the other 16 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index3 boards, none shared
Math benchmarks for DeepSeek V4 Flash 0423 and Mercury 2
BenchmarkDeepSeek V4 Flash 0423Mercury 2Leader
LMArena Text: math1426Not measuredOnly DeepSeek V4 Flash 0423 measured
LMArena Text: mathematical work1432Not measuredOnly DeepSeek V4 Flash 0423 measured
MathArena24.8%Not measuredOnly DeepSeek V4 Flash 0423 measured
Neither model has a result on the other 11 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index30
Knowledge benchmarks for DeepSeek V4 Flash 0423 and Mercury 2
BenchmarkDeepSeek V4 Flash 0423Mercury 2Leader
LMArena Text: science14461343DeepSeek V4 Flash 0423
LMArena Text: medicine and healthcare14481363DeepSeek V4 Flash 0423
LMArena Text: legal and government1439Not measuredOnly DeepSeek V4 Flash 0423 measured
LMArena Text: business and finance14251349DeepSeek V4 Flash 0423
Neither model has a result on the other 16 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index80
Web design benchmarks for DeepSeek V4 Flash 0423 and Mercury 2
BenchmarkDeepSeek V4 Flash 0423Mercury 2Leader
Design Arena: websites12071035DeepSeek V4 Flash 0423
Design Arena: UI components1162996DeepSeek V4 Flash 0423
Design Arena: game dev1196983DeepSeek V4 Flash 0423
Design Arena: data viz11271010DeepSeek V4 Flash 0423
Design Arena: 3D1188982DeepSeek V4 Flash 0423
Design Arena: all code categories12051021DeepSeek V4 Flash 0423
Design Arena: SVG1154997DeepSeek V4 Flash 0423
Design Arena: ASCII art1109988DeepSeek V4 Flash 0423
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index110
General benchmarks for DeepSeek V4 Flash 0423 and Mercury 2
BenchmarkDeepSeek V4 Flash 0423Mercury 2Leader
IFBenchNot measured69.8%Only Mercury 2 measured
Long context reasoning (AA-LCR)Not measured43.7%Only Mercury 2 measured
DTBench86.4%Not measuredOnly DeepSeek V4 Flash 0423 measured
LMCA35.9%Not measuredOnly DeepSeek V4 Flash 0423 measured
LMArena Text14321354DeepSeek V4 Flash 0423
LMArena Text: instruction following14211328DeepSeek V4 Flash 0423
LMArena Text: creative writing14031286DeepSeek V4 Flash 0423
LMArena Text: English14411378DeepSeek V4 Flash 0423
LMArena Text: non-English14191330DeepSeek V4 Flash 0423
LMArena Text: Chinese14631414DeepSeek V4 Flash 0423
LMArena Text: German1419Not measuredOnly DeepSeek V4 Flash 0423 measured
LMArena Text: Japanese1389Not measuredOnly DeepSeek V4 Flash 0423 measured
LMArena Text: Korean1383Not measuredOnly DeepSeek V4 Flash 0423 measured
LMArena Text: French1439Not measuredOnly DeepSeek V4 Flash 0423 measured
LMArena Text: Spanish1436Not measuredOnly DeepSeek V4 Flash 0423 measured
LMArena Text: Russian14271304DeepSeek V4 Flash 0423
LMArena Text: Polish1436Not measuredOnly DeepSeek V4 Flash 0423 measured
LMArena Text: multi-turn14491357DeepSeek V4 Flash 0423
LMArena Text: longer queries14341329DeepSeek V4 Flash 0423
LMArena Text: writing and language14101300DeepSeek V4 Flash 0423
LMArena Text: entertainment and media14001314DeepSeek V4 Flash 0423
Neither model has a result on the other 27 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for DeepSeek V4 Flash 0423 and Mercury 2
BenchmarkDeepSeek V4 Flash 0423Mercury 2Leader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

DeepSeek V4 Flash 0423$0.0003Fits in one request
Mercury 2$0.0006Fits in one request

DeepSeek V4 Flash 0423 costs less

Repository review

50K input + 3K output tokens

DeepSeek V4 Flash 0423$0.0073Fits in one request
Mercury 2$0.015Fits in one request

DeepSeek V4 Flash 0423 costs less

Agent loop

200K input + 10K output tokens

DeepSeek V4 Flash 0423$0.029Fits in one request
Mercury 2$0.058Exceeds the context window

DeepSeek V4 Flash 0423 costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of DeepSeek V4 Flash 0423 and Mercury 2
SpecificationDeepSeek V4 Flash 0423Inception Mercury 2
MakerDeepSeekInception
WeightsOpen weightsNo public weights listed
ReleasedApril 24, 2026March 4, 2026
Context window1.05M128K
Input $/Mtok$0.13$0.25
Output $/Mtok$0.28$0.75
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it11
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is DeepSeek V4 Flash 0423 better than Mercury 2 for coding?

Neither model has a coding score yet, so the two cannot be ranked on coding. Pick DeepSeek V4 Flash 0423 for output tokens 2.7x cheaper ($0.28 vs $0.75 per million) and a larger context window (1.05M against 128K tokens).

Which is cheaper, DeepSeek V4 Flash 0423 or Mercury 2?

DeepSeek V4 Flash 0423 costs $0.13 per million input tokens and $0.28 per million output tokens; Mercury 2 costs $0.25 and $0.75. Cost per solved task is known for neither of them.

Which has the bigger context window, DeepSeek V4 Flash 0423 or Mercury 2?

DeepSeek V4 Flash 0423, with 1.05M tokens against 128K.

How many benchmarks is the DeepSeek V4 Flash 0423 vs Mercury 2 comparison based on?

35 benchmarks have a result for both models as of October 22, 2026. DeepSeek V4 Flash 0423's coding index rests on 1 independent benchmark and Mercury 2's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next