Enter opens the highlighted result. Esc closes.

DeepSeek V3.2 vs Gemini 4 Neon

Gemini 4 Neon leads DeepSeek V3.2 on coding, 86.9 to 30.0 on the Vibe Coding Index, and DeepSeek V3.2 costs 24.6x less per output token ($0.61 vs $15.00 per million). Updated November 22, 2026; 35 benchmarks have a result for both.

DeepSeek V3.2DeepSeek
Coding
30.0
Overall
40.0
Output $/M
$0.61
0
Gemini 4 NeonGoogle
Coding
86.9
Overall
89.1
Output $/M
$15.00
35
vs boards won of 35 shared

DeepSeek V3.2 or Gemini 4 Neon: which to pick

Pick Gemini 4 Neon for coding: it scores 86.9 on the Vibe Coding Index against 30.0 for DeepSeek V3.2.

Pick DeepSeek V3.2 for output tokens 24.6x cheaper ($0.61 vs $15.00 per million).

Shared results
35
DeepSeek V3.2 only
40
Gemini 4 Neon only
34
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 V3.2 30.0 from 3 benchmarks, Gemini 4 Neon 86.9 from 4 benchmarks.
Widest category
Agentic work, 77.6 points apart on average over 2 shared benchmarks.
Widest single gap
Vending-Bench 2: DeepSeek V3.2 $1,034 against Gemini 4 Neon $16,843, 93.5 points apart on a 0 to 100 scale.
Output price
DeepSeek V3.2 $0.61 against Gemini 4 Neon $15.00 per million output tokens, 24.6x apart.
Context window
DeepSeek V3.2 takes 164K tokens, Gemini 4 Neon 1.05M.

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 V3.2BarsGemini 4 NeonTies
Coding06
Agentic work02
Reasoning03
Math02
Knowledge04
Web design05
General013
All shared boards035
CodingLike-for-likeSix of these boards feed the Vibe Coding Index06
Coding benchmarks for DeepSeek V3.2 and Gemini 4 Neon
BenchmarkDeepSeek V3.2Gemini 4 NeonLeader
SWE-bench Verified70.0%Not measuredOnly DeepSeek V3.2 measured
SWE-bench Multilingual59.0%Not measuredOnly DeepSeek V3.2 measured
LiveBench codingNot measured89.7%Only Gemini 4 Neon measured
LMArena WebDev13621765Gemini 4 Neon
LiveCodeBench (Vals AI)80.7%Not measuredOnly DeepSeek V3.2 measured
SWE-bench Verified (Vals AI)67.6%Not measuredOnly DeepSeek V3.2 measured
Terminal-Bench 4 (Vals AI)Not measured61.6%Only Gemini 4 Neon measured
Vibe Code Bench (Vals AI)5.1%93.7%Gemini 4 Neon
LiveCodeBench (Artificial Analysis run)86.2%Not measuredOnly DeepSeek V3.2 measured
Terminal-Bench Hard (Artificial Analysis run)35.6%Not measuredOnly DeepSeek V3.2 measured
Terminal-Bench 4.0 (Artificial Analysis run)Not measured63.6%Only Gemini 4 Neon measured
FrontierCode 2.0Not measured44.4%Only Gemini 4 Neon measured
Terminal-Bench39.6%Not measuredOnly DeepSeek V3.2 measured
LMArena Text: coding14541555Gemini 4 Neon
LMArena Text: software and IT14501542Gemini 4 Neon
LMArena WebDev: React13491787Gemini 4 Neon
LMArena WebDev: HTML13701803Gemini 4 Neon
GameBench 2Not measured98.0Only Gemini 4 Neon measured
Terminal-Bench 2.039.6%Not measuredOnly DeepSeek V3.2 measured
CursorBench 4.0Not measured65.4%Only Gemini 4 Neon measured
Neither model has a result on the other 32 boards here
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
Intelligence benchmarks for DeepSeek V3.2 and Gemini 4 Neon
BenchmarkDeepSeek V3.2Gemini 4 NeonLeader
AA Intelligence IndexNot measured61.4Only Gemini 4 Neon measured
Epoch Capabilities Index146.3Not measuredOnly DeepSeek V3.2 measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index02
Agentic work benchmarks for DeepSeek V3.2 and Gemini 4 Neon
BenchmarkDeepSeek V3.2Gemini 4 NeonLeader
LiveBench researchNot measured94.5%Only Gemini 4 Neon measured
LiveBench browserNot measured95.6%Only Gemini 4 Neon measured
LiveBench long agenticNot measured82.6%Only Gemini 4 Neon measured
Tax Agent Bench (Vals AI)Not measured54.9%Only Gemini 4 Neon measured
Legal Research Bench (Vals AI)Not measured56.3%Only Gemini 4 Neon measured
Harvey's Legal Agent Benchmark (Vals AI)Not measured19.6%Only Gemini 4 Neon measured
BioMysteryBench (Vals AI)Not measured80.0%Only Gemini 4 Neon measured
Terminal-Bench-Science (Vals AI)Not measured62.9%Only Gemini 4 Neon measured
Vals Science IndexNot measured62.6%Only Gemini 4 Neon measured
τ²-Bench Telecom90.6%Not measuredOnly DeepSeek V3.2 measured
APEX-Agents21.3%82.9%Gemini 4 Neon
Vending-Bench 2$1,034$16,843Gemini 4 Neon
Neither model has a result on the other 32 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index03
Reasoning benchmarks for DeepSeek V3.2 and Gemini 4 Neon
BenchmarkDeepSeek V3.2Gemini 4 NeonLeader
GPQA Diamond80.3%Not measuredOnly DeepSeek V3.2 measured
LiveBench reasoningNot measured89.6%Only Gemini 4 Neon measured
MysteryMechanism (Vals AI)Not measured45.0%Only Gemini 4 Neon measured
GPQA Diamond (Artificial Analysis run)84.0%Not measuredOnly DeepSeek V3.2 measured
Humanity's Last Exam (Artificial Analysis run)24.6%Not measuredOnly DeepSeek V3.2 measured
HLE-Diamond (Artificial Analysis run)Not measured71.3%Only Gemini 4 Neon measured
CritPt v2, hard subset (Artificial Analysis run)Not measured61.5%Only Gemini 4 Neon measured
GPQA Diamond (Epoch AI run)Not measured96.3%Only Gemini 4 Neon measured
ARC-AGI57.0%Not measuredOnly DeepSeek V3.2 measured
ARC-AGI-24.0%Not measuredOnly DeepSeek V3.2 measured
SimpleBenchNot measured94.4%Only Gemini 4 Neon measured
Chess puzzlesNot measured83.0%Only Gemini 4 Neon measured
Mystery game puzzlesNot measured85.0%Only Gemini 4 Neon measured
EBR-benchNot measured98.1%Only Gemini 4 Neon measured
LMArena Text: hard prompts14341547Gemini 4 Neon
LMArena Text: expert14351560Gemini 4 Neon
LMArena Text: hard prompts in English14461524Gemini 4 Neon
Neither model has a result on the other 9 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index02
Math benchmarks for DeepSeek V3.2 and Gemini 4 Neon
BenchmarkDeepSeek V3.2Gemini 4 NeonLeader
AIME84.6%Not measuredOnly DeepSeek V3.2 measured
MGSM (Vals AI)86.0%Not measuredOnly DeepSeek V3.2 measured
AIME 2025 (Artificial Analysis run)92.0%Not measuredOnly DeepSeek V3.2 measured
OTIS Mock AIMENot measured100.0%Only Gemini 4 Neon measured
FrontierMath, tiers 1 to 3Not measured91.2%Only Gemini 4 Neon measured
FrontierMath, tier 4Not measured97.6%Only Gemini 4 Neon measured
LMArena Text: math14351548Gemini 4 Neon
LMArena Text: mathematical work14381566Gemini 4 Neon
MathArena14.4%Not measuredOnly DeepSeek V3.2 measured
Neither model has a result on the other 5 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
Knowledge benchmarks for DeepSeek V3.2 and Gemini 4 Neon
BenchmarkDeepSeek V3.2Gemini 4 NeonLeader
MMLU Pro84.9%Not measuredOnly DeepSeek V3.2 measured
LegalBench (Vals AI)76.1%Not measuredOnly DeepSeek V3.2 measured
CaseLaw (Vals AI)55.4%Not measuredOnly DeepSeek V3.2 measured
CorpFin (Vals AI)51.0%Not measuredOnly DeepSeek V3.2 measured
TaxEval (Vals AI)68.2%Not measuredOnly DeepSeek V3.2 measured
MedQA (Vals AI)93.9%Not measuredOnly DeepSeek V3.2 measured
MMLU Pro (Artificial Analysis run)86.2%Not measuredOnly DeepSeek V3.2 measured
AA-OmniscienceNot measured54.8Only Gemini 4 Neon measured
SimpleQA VerifiedNot measured75.2%Only Gemini 4 Neon measured
LMArena Text: science14461549Gemini 4 Neon
LMArena Text: medicine and healthcare14401534Gemini 4 Neon
LMArena Text: legal and government14251518Gemini 4 Neon
LMArena Text: business and finance14101493Gemini 4 Neon
Neither model has a result on the other 8 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index05
Web design benchmarks for DeepSeek V3.2 and Gemini 4 Neon
BenchmarkDeepSeek V3.2Gemini 4 NeonLeader
Design Arena: websites11621347Gemini 4 Neon
Design Arena: UI components11321349Gemini 4 Neon
Design Arena: game dev11031394Gemini 4 Neon
Design Arena: data viz1153Not measuredOnly DeepSeek V3.2 measured
Design Arena: 3D10991490Gemini 4 Neon
Design Arena: all code categories11461391Gemini 4 Neon
Design Arena: SVG1016Not measuredOnly DeepSeek V3.2 measured
Design Arena: ASCII art1068Not measuredOnly DeepSeek V3.2 measured
LiveBench designNot measured81.5%Only Gemini 4 Neon measured
GeneralLike-for-likePublic boards, shown as published, kept out of the index013
General benchmarks for DeepSeek V3.2 and Gemini 4 Neon
BenchmarkDeepSeek V3.2Gemini 4 NeonLeader
Vals IndexNot measured70.9%Only Gemini 4 Neon measured
LiveBench data analysisNot measured85.3%Only Gemini 4 Neon measured
IFBench60.7%Not measuredOnly DeepSeek V3.2 measured
Long context reasoning (AA-LCR)73.3%Not measuredOnly DeepSeek V3.2 measured
GDPval-AA698Not measuredOnly DeepSeek V3.2 measured
DTBench85.6%Not measuredOnly DeepSeek V3.2 measured
LMCA28.8%Not measuredOnly DeepSeek V3.2 measured
CL-bench12.4%Not measuredOnly DeepSeek V3.2 measured
Furniture assemblyNot measured89.2%Only Gemini 4 Neon measured
CL-bench Life7.4%Not measuredOnly DeepSeek V3.2 measured
LMArena Text14251524Gemini 4 Neon
LMArena Text: instruction following14131535Gemini 4 Neon
LMArena Text: creative writing14001519Gemini 4 Neon
LMArena Text: English14391511Gemini 4 Neon
LMArena Text: non-English14091524Gemini 4 Neon
LMArena Text: Chinese14551598Gemini 4 Neon
LMArena Text: German1412Not measuredOnly DeepSeek V3.2 measured
LMArena Text: Japanese1379Not measuredOnly DeepSeek V3.2 measured
LMArena Text: Korean1370Not measuredOnly DeepSeek V3.2 measured
LMArena Text: French14301563Gemini 4 Neon
LMArena Text: Spanish14251505Gemini 4 Neon
LMArena Text: Russian14231554Gemini 4 Neon
LMArena Text: Polish1419Not measuredOnly DeepSeek V3.2 measured
LMArena Text: multi-turn14271525Gemini 4 Neon
LMArena Text: longer queries14281533Gemini 4 Neon
LMArena Text: writing and language14071517Gemini 4 Neon
LMArena Text: entertainment and media13951495Gemini 4 Neon
Neither model has a result on the other 22 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for DeepSeek V3.2 and Gemini 4 Neon
BenchmarkDeepSeek V3.2Gemini 4 NeonLeader
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 V3.2$0.0006Fits in one request
Gemini 4 Neon$0.011Fits in one request

DeepSeek V3.2 costs less

Repository review

50K input + 3K output tokens

DeepSeek V3.2$0.015Fits in one request
Gemini 4 Neon$0.195Fits in one request

DeepSeek V3.2 costs less

Agent loop

200K input + 10K output tokens

DeepSeek V3.2$0.060Exceeds the context window
Gemini 4 Neon$0.750Fits in one request

DeepSeek V3.2 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 V3.2 and Gemini 4 Neon
SpecificationDeepSeek V3.2Google Gemini 4 Neon
MakerDeepSeekGoogle
WeightsOpen weightsNo public weights listed
ReleasedDecember 1, 2025November 18, 2026
Context window164K1.05M
Input $/Mtok$0.27$3.00
Output $/Mtok$0.61$15.00
Coding index30.086.9
Independent measurements behind it34
Cost per solved task$0.99Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is DeepSeek V3.2 better than Gemini 4 Neon for coding?

Pick Gemini 4 Neon for coding: it scores 86.9 on the Vibe Coding Index against 30.0 for DeepSeek V3.2. Pick DeepSeek V3.2 for output tokens 24.6x cheaper ($0.61 vs $15.00 per million).

Which is cheaper, DeepSeek V3.2 or Gemini 4 Neon?

DeepSeek V3.2 costs $0.27 per million input tokens and $0.61 per million output tokens; Gemini 4 Neon costs $3.00 and $15.00. Cost per solved task is known for DeepSeek V3.2 only.

Which has the bigger context window, DeepSeek V3.2 or Gemini 4 Neon?

Gemini 4 Neon, with 1.05M tokens against 164K.

How many benchmarks is the DeepSeek V3.2 vs Gemini 4 Neon comparison based on?

35 benchmarks have a result for both models as of November 22, 2026. DeepSeek V3.2's coding index rests on 3 independent benchmarks and Gemini 4 Neon's on 4. 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-v3-2-vs-gemini-4-neon" title="DeepSeek V3.2 vs Gemini 4 Neon, 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-v3-2-vs-gemini-4-neon">LLMBenchmark.ai, DeepSeek V3.2 vs Gemini 4 Neon</a></p>