Enter opens the highlighted result. Esc closes.

DeepSeek V4 Pro 0423 vs Grok 4.20

DeepSeek V4 Pro 0423 leads Grok 4.20 on coding, 38.7 to 33.3 on the Vibe Coding Index, at about the same output price ($2.716 vs $2.50 per million). Updated October 21, 2026; 36 benchmarks have a result for both.

Replace DeepSeek V4 Pro 0423 Replace Grok 4.20
DeepSeek V4 Pro 0423DeepSeek
Coding
38.7
Overall
50.0
Output $/M
$2.716
24
Grok 4.20xAI
Coding
33.3
Overall
41.0
Output $/M
$2.50
12
vs boards won of 36 shared

DeepSeek V4 Pro 0423 or Grok 4.20: which to pick

Pick DeepSeek V4 Pro 0423 for coding: it scores 38.7 on the Vibe Coding Index against 33.3 for Grok 4.20.

Pick Grok 4.20 for output tokens 1.1x cheaper ($2.50 vs $2.716 per million) and a larger context window (2M against 1.05M tokens).

Model ADeepSeek V4 Pro 0423DeepSeek38.7/1005 independent benchmarks$0.57 per solved taskModel BGrok 4.20xAI33.3/1002 independent benchmarks$0.72 per solved task
Shared results
36
DeepSeek V4 Pro 0423 only
48
Grok 4.20 only
19
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 Pro 0423 38.7 from 5 benchmarks, Grok 4.20 33.3 from 2 benchmarks.
Widest category
Agentic work, 15.8 points apart on average over 8 shared benchmarks.
Widest single gap
Vibe Code Bench (Vals AI): DeepSeek V4 Pro 0423 49.9% against Grok 4.20 4.1%, 45.9 points apart on a 0 to 100 scale.
Output price
DeepSeek V4 Pro 0423 $2.716 against Grok 4.20 $2.50 per million output tokens, 1.1x apart.
Context window
DeepSeek V4 Pro 0423 takes 1.05M tokens, Grok 4.20 2M.

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 Pro 0423BarsGrok 4.20Ties
Coding51
Intelligence11
Agentic work44
Reasoning11
Knowledge42
Web design53
General40
All shared boards2412
CodingLike-for-likeSix of these boards feed the Vibe Coding Index51
Coding benchmarks for DeepSeek V4 Pro 0423 and Grok 4.20
BenchmarkDeepSeek V4 Pro 0423Grok 4.20Leader
LMArena WebDev1447Not measuredOnly DeepSeek V4 Pro 0423 measured
AA Coding Index59.4Not measuredOnly DeepSeek V4 Pro 0423 measured
LiveCodeBench (Vals AI)87.5%84.3%DeepSeek V4 Pro 0423
SWE-bench Verified (Vals AI)77.4%72.2%DeepSeek V4 Pro 0423
Terminal-Bench 4 (Vals AI)11.1%Not measuredOnly DeepSeek V4 Pro 0423 measured
Vibe Code Bench (Vals AI)49.9%4.1%DeepSeek V4 Pro 0423
Code migration (Vals AI)26.2%0.3%DeepSeek V4 Pro 0423
ProgramBench (Vals AI)0.0%Not measuredOnly DeepSeek V4 Pro 0423 measured
Terminal-Bench 2.1 (Vals AI)50.2%44.2%DeepSeek V4 Pro 0423
SkillsBench (Vals AI)51.3%Not measuredOnly DeepSeek V4 Pro 0423 measured
Terminal-Bench Hard (Artificial Analysis run)Not measured37.9%Only Grok 4.20 measured
Terminal-BenchNot measured57.3%Only Grok 4.20 measured
FrontierCode17.6%Not measuredOnly DeepSeek V4 Pro 0423 measured
WeirdML48.9%52.3%Grok 4.20
ALE-Bench1006.1Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: coding1470Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: software and IT1469Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena WebDev: React1441Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena WebDev: HTML1435Not measuredOnly DeepSeek V4 Pro 0423 measured
Neither model has a result on the other 20 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index11
Intelligence benchmarks for DeepSeek V4 Pro 0423 and Grok 4.20
BenchmarkDeepSeek V4 Pro 0423Grok 4.20Leader
AA Intelligence Index30.425.7DeepSeek V4 Pro 0423
Epoch Capabilities Index149.1152.1Grok 4.20
Agentic workLike-for-likePublic boards, shown as published, kept out of the index44
Agentic work benchmarks for DeepSeek V4 Pro 0423 and Grok 4.20
BenchmarkDeepSeek V4 Pro 0423Grok 4.20Leader
AA Agentic Index26.3Not measuredOnly DeepSeek V4 Pro 0423 measured
Design Arena: agents building web apps10001144Grok 4.20
Design Arena: agents building full-stack apps9481054Grok 4.20
Design Arena: agents building mobile appsNot measured1090Only Grok 4.20 measured
Design Arena: agents building Android appsNot measured1052Only Grok 4.20 measured
Design Arena: agents building Godot games10591089Grok 4.20
Design Arena: HTML slidesNot measured1180Only Grok 4.20 measured
Finance Agent (Vals AI)44.1%28.5%DeepSeek V4 Pro 0423
Tax Agent Bench (Vals AI)25.2%7.0%DeepSeek V4 Pro 0423
Legal Research Bench (Vals AI)23.1%13.9%DeepSeek V4 Pro 0423
Excel Modeling Benchmark (Vals AI)51.6%11.7%DeepSeek V4 Pro 0423
τ²-Bench TelecomNot measured93.0%Only Grok 4.20 measured
Vending-Bench 2$3,285$4,663Grok 4.20
Neither model has a result on the other 28 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index11
Reasoning benchmarks for DeepSeek V4 Pro 0423 and Grok 4.20
BenchmarkDeepSeek V4 Pro 0423Grok 4.20Leader
GPQA Diamond89.4%88.6%DeepSeek V4 Pro 0423
MMMU Pro (Vals AI)Not measured83.5%Only Grok 4.20 measured
GPQA Diamond (Artificial Analysis run)Not measured91.1%Only Grok 4.20 measured
Humanity's Last Exam (Artificial Analysis run)Not measured34.5%Only Grok 4.20 measured
GPQA Diamond (Epoch AI run)90.9%Not measuredOnly DeepSeek V4 Pro 0423 measured
ARC-AGINot measured89.5%Only Grok 4.20 measured
ARC-AGI-2Not measured65.1%Only Grok 4.20 measured
Chess puzzles19.0%Not measuredOnly DeepSeek V4 Pro 0423 measured
Mystery game puzzles17.0%Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: hard prompts1461Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: expert1463Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: hard prompts in English1466Not measuredOnly DeepSeek V4 Pro 0423 measured
Kagi LLM Benchmark53.5%75.0%Grok 4.20
Neither model has a result on the other 9 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
Math benchmarks for DeepSeek V4 Pro 0423 and Grok 4.20
BenchmarkDeepSeek V4 Pro 0423Grok 4.20Leader
AIMENot measured96.5%Only Grok 4.20 measured
OTIS Mock AIME96.7%Not measuredOnly DeepSeek V4 Pro 0423 measured
FrontierMath, tiers 1 to 345.3%Not measuredOnly DeepSeek V4 Pro 0423 measured
FrontierMath, tier 42.4%Not measuredOnly DeepSeek V4 Pro 0423 measured
ProofBench16.0%Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: math1443Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: mathematical work1455Not measuredOnly DeepSeek V4 Pro 0423 measured
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index42
Knowledge benchmarks for DeepSeek V4 Pro 0423 and Grok 4.20
BenchmarkDeepSeek V4 Pro 0423Grok 4.20Leader
MMLU Pro87.3%86.3%DeepSeek V4 Pro 0423
LegalBench (Vals AI)80.3%77.7%DeepSeek V4 Pro 0423
CaseLaw (Vals AI)59.4%54.4%DeepSeek V4 Pro 0423
CorpFin (Vals AI)61.4%63.7%Grok 4.20
TaxEval (Vals AI)72.1%74.1%Grok 4.20
MortgageTax (Vals AI)Not measured45.4%Only Grok 4.20 measured
MedQA (Vals AI)Not measured94.5%Only Grok 4.20 measured
MedCode (Vals AI)40.5%32.2%DeepSeek V4 Pro 0423
SimpleQA Verified47.0%Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: science1470Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: medicine and healthcare1466Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: legal and government1464Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: business and finance1443Not measuredOnly DeepSeek V4 Pro 0423 measured
Neither model has a result on the other 5 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index53
Web design benchmarks for DeepSeek V4 Pro 0423 and Grok 4.20
BenchmarkDeepSeek V4 Pro 0423Grok 4.20Leader
Design Arena: websites12341229DeepSeek V4 Pro 0423
Design Arena: UI components12221199DeepSeek V4 Pro 0423
Design Arena: game dev12291192DeepSeek V4 Pro 0423
Design Arena: data viz12031217Grok 4.20
Design Arena: 3D12541193DeepSeek V4 Pro 0423
Design Arena: all code categories12401221DeepSeek V4 Pro 0423
Design Arena: SVG11401164Grok 4.20
Design Arena: ASCII art11491174Grok 4.20
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index40
General benchmarks for DeepSeek V4 Pro 0423 and Grok 4.20
BenchmarkDeepSeek V4 Pro 0423Grok 4.20Leader
Public Benefits Bench (Vals AI)62.9%Not measuredOnly DeepSeek V4 Pro 0423 measured
SAGE (Vals AI)Not measured38.2%Only Grok 4.20 measured
MedScribe (Vals AI)75.1%63.4%DeepSeek V4 Pro 0423
Vals Multimodal IndexNot measured39.1%Only Grok 4.20 measured
IFBenchNot measured81.2%Only Grok 4.20 measured
Long context reasoning (AA-LCR)Not measured69.0%Only Grok 4.20 measured
DTBench90.7%90.1%DeepSeek V4 Pro 0423
LMCA41.2%38.7%DeepSeek V4 Pro 0423
CL-benchNot measured22.2%Only Grok 4.20 measured
Surface Evolver Bench40.0%Not measuredOnly DeepSeek V4 Pro 0423 measured
CL-bench Life13.5%11.9%DeepSeek V4 Pro 0423
LMArena Text1451Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: instruction following1445Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: creative writing1446Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: English1459Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: non-English1439Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: Chinese1481Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: German1458Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: Japanese1446Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: Korean1409Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: French1473Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: Spanish1457Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: Russian1453Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: Polish1461Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: multi-turn1468Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: longer queries1458Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: writing and language1447Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: entertainment and media1431Not measuredOnly DeepSeek V4 Pro 0423 measured
Neither model has a result on the other 18 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for DeepSeek V4 Pro 0423 and Grok 4.20
BenchmarkDeepSeek V4 Pro 0423Grok 4.20Leader
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 Pro 0423$0.0026Fits in one request
Grok 4.20$0.0025Fits in one request

Grok 4.20 costs less

Repository review

50K input + 3K output tokens

DeepSeek V4 Pro 0423$0.069Fits in one request
Grok 4.20$0.070Fits in one request

DeepSeek V4 Pro 0423 costs less

Agent loop

200K input + 10K output tokens

DeepSeek V4 Pro 0423$0.272Fits in one request
Grok 4.20$0.275Fits in one request

DeepSeek V4 Pro 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 Pro 0423 and Grok 4.20
SpecificationDeepSeek V4 Pro 0423SpaceXAI Grok 4.20
MakerDeepSeekxAI
WeightsOpen weightsNo public weights listed
ReleasedApril 24, 2026March 31, 2026
Context window1.05M2M
Input $/Mtok$1.225$1.25
Output $/Mtok$2.716$2.50
Coding index38.733.3
Independent measurements behind it52
Cost per solved task$0.57$0.72
SourceOpenRouter listingOpenRouter listing

Questions

Is DeepSeek V4 Pro 0423 better than Grok 4.20 for coding?

Pick DeepSeek V4 Pro 0423 for coding: it scores 38.7 on the Vibe Coding Index against 33.3 for Grok 4.20. Pick Grok 4.20 for output tokens 1.1x cheaper ($2.50 vs $2.716 per million) and a larger context window (2M against 1.05M tokens).

Which is cheaper, DeepSeek V4 Pro 0423 or Grok 4.20?

DeepSeek V4 Pro 0423 costs $1.225 per million input tokens and $2.716 per million output tokens; Grok 4.20 costs $1.25 and $2.50. Per solved SWE-bench Verified task, DeepSeek V4 Pro 0423 costs $0.57 and Grok 4.20 $0.72.

Which has the bigger context window, DeepSeek V4 Pro 0423 or Grok 4.20?

Grok 4.20, with 2M tokens against 1.05M.

How many benchmarks is the DeepSeek V4 Pro 0423 vs Grok 4.20 comparison based on?

36 benchmarks have a result for both models as of October 21, 2026. DeepSeek V4 Pro 0423's coding index rests on 5 independent benchmarks and Grok 4.20's on 2. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/deepseek-v4-pro-vs-grok-4-20" title="DeepSeek V4 Pro 0423 vs Grok 4.20, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/deepseek-v4-pro-vs-grok-4-20">LLMBenchmark.ai, DeepSeek V4 Pro 0423 vs Grok 4.20</a></p>