BenchLeader

DeepSeek V4 Pro vs Gemini 3.7 Flash

Full profiles: DeepSeek V4 Pro (DeepSeek) and Gemini 3.7 Flash (Google): every benchmark, price and rank.

Updated · Data from 11 sources

As of 16 Nov 2026, Gemini 3.7 Flash is ahead of DeepSeek V4 Pro on quality (62.1 vs 56.0 on the BenchLeader Index), but DeepSeek V4 Pro is 1.5× cheaper at $0.990 vs $1.50 per 1M tokens blended. Choose DeepSeek V4 Pro for maths (65 vs 63); choose Gemini 3.7 Flash for fast first answers (2.2 s vs 42.8 s) and knowledge (66 vs 41), reasoning (68 vs 56), and coding (62 vs 52), where it also streams ~2× faster.

Key facts (as of 16 Nov 2026)

  • DeepSeek V4 Pro scores 56.0 on the BenchLeader Index and Gemini 3.7 Flash scores 62.1 (BenchLeader, 16 Nov 2026).
  • DeepSeek V4 Pro costs $0.990 per 1M tokens blended (list price, 16 Nov 2026).
  • Gemini 3.7 Flash costs $1.50 per 1M tokens blended (list price, 16 Nov 2026).
  • Gemini 3.7 Flash streams ~2× faster than DeepSeek V4 Pro (Artificial Analysis, 7-day median, 16 Nov 2026).
  • Of the 25 benchmarks both models have, Gemini 3.7 Flash leads 22 (BenchLeader, 16 Nov 2026).
  • On Arena WebDev, DeepSeek V4 Pro is rated 1581 and Gemini 3.7 Flash 1588 (LMArena, as published by 16 Nov 2026).

Which to pick for…

The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.

Use casePickWhyConfidence
CodingGemini 3.7 Flash62 vs 52 category score; Arena WebDev 1588 vs 1581High
MathsDeepSeek V4 Pro65 vs 63 category score; Arena Maths 1465 vs 1502Medium
ReasoningGemini 3.7 Flash68 vs 56 category score; GPQA Diamond 94.8% vs 90.9%High
KnowledgeGemini 3.7 Flash66 vs 41 category scoreHigh
Long documentsGemini 3.7 Flash66 vs 64 category score; AA-LCR v1.1 81.7% vs 79.3%Medium
Chat quality (human preference)Gemini 3.7 Flash68 vs 65 on Arena votesMedium
Lowest costGemini 3.7 Flash$1.50 vs $1.98 per 1M blended; agentic coding session $0.060 vs $0.095–
Fastest first answerGemini 3.7 Flash2.2 s (Gemini 3.7 Flash (high)) vs 42.8 s (DeepSeek V4 Pro (high)) to the first answer token–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • Gemini 3.7 Flash (high) leads on quality: 62.1 vs 56.0.
  • DeepSeek V4 Pro (high) is stronger in maths.
  • Gemini 3.7 Flash (high) is stronger in coding, human preference, knowledge, long context, reasoning.
  • DeepSeek V4 Pro (high) is 1.5× cheaper ($0.990 vs $1.50 per 1M blended).

At a glance

SpecDeepSeek V4 ProGemini 3.7 Flash
DeveloperDeepSeekGoogle
Released13 Aug 202613 Aug 2026
Context window (tokens)1,048,5761,048,576
Compared atDeepSeek V4 Pro (high)Gemini 3.7 Flash (high)
Open weightsYesNo
Price in / out / cached input per 1M$1.32 / $3.96 / $0.022$0.750 / $3.75 / $0.075

Category by category

Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.

CategoryDeepSeek V4 Pro (high)Gemini 3.7 Flash (high)Leader
Reasoning55.667.8Gemini 3.7 Flash
Coding51.562.3Gemini 3.7 Flash
Agents & tools–48.9not comparable
Maths64.562.6DeepSeek V4 Pro
Knowledge41.265.9Gemini 3.7 Flash
Human preference65.068.1Gemini 3.7 Flash
Multimodal–67.2not comparable
Long context64.365.5Gemini 3.7 Flash

On one chart

Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.

2550 avg75ReasoningCodingMathsKnowledgeLong contextHumanpreferenceDeepSeek V4 Pro, Reasoning: 55.6DeepSeek V4 Pro, Coding: 51.5DeepSeek V4 Pro, Maths: 64.5DeepSeek V4 Pro, Knowledge: 41.2DeepSeek V4 Pro, Long context: 64.3DeepSeek V4 Pro, Human preference: 65.0Gemini 3.7 Flash, Reasoning: 67.8Gemini 3.7 Flash, Coding: 62.3Gemini 3.7 Flash, Maths: 62.6Gemini 3.7 Flash, Knowledge: 65.9Gemini 3.7 Flash, Long context: 65.5Gemini 3.7 Flash, Human preference: 68.1

Not drawn, measured for fewer than two of these models: Agents & tools, Instruction following, Multimodal.

Model, click to add or removeIndex

Where sources disagree

Across the 25 benchmarks both have, DeepSeek V4 Pro leads 3, Gemini 3.7 Flash leads 22. Every overall index that rates both puts Gemini 3.7 Flash ahead.

Overall indexDeepSeek V4 Pro (high)Gemini 3.7 Flash (high)Leader
BenchLeader Index56.062.1Gemini 3.7 Flash
Arena Text14641489Gemini 3.7 Flash

Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).

Price, speed and latency

List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.

MeasureDeepSeek V4 Pro (high)Gemini 3.7 Flash (high)
Blended price per 1M tokens$1.98$1.50
Input / output per 1M$1.32 / $3.96$0.750 / $3.75
Cached input per 1M$0.022$0.075
Price per task: chat reply$0.0017$0.0014
Price per task: summarize a 30-page report$0.018$0.011
Price per task: code edit$0.014$0.010
Price per task: agentic coding session$0.095$0.060
Price per task: structured extraction$0.0034$0.0023
Agentic coding session, input mostly cached$0.037$0.030
Cost per benchmark task (AA Index)$0.849–
Output tokens per task (AA Index)70k–
Of which reasoning45k–
Time per task (AA Index)15 min–
Output speed80 tok/s95 tok/s
Time to first answer42.8 s2.2 s

Each model is compared at its best-ranked reasoning-effort setting. DeepSeek V4 Pro (high) scores 56.0 and takes 43 s to its first answer token; DeepSeek V4 Pro (max) scores 54.7 in 47 s, DeepSeek V4 Pro scores 53.0 in 1 s. Gemini 3.7 Flash (high) scores 62.1 and takes 2 s to its first answer token; Gemini 3.7 Flash (medium) scores 60.1 in 2 s, Gemini 3.7 Flash (low) scores 57.3 in 2 s. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.

Every benchmark

140 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.

MetricPublisherDeepSeek V4 Pro (high)Gemini 3.7 Flash (high)
BenchLeader IndexBenchLeader56.062.1
Coding scoreBenchLeader51.562.3
Human preference scoreBenchLeader65.068.1
Knowledge scoreBenchLeader41.265.9
Long context scoreBenchLeader64.365.5
Maths scoreBenchLeader64.562.6
Reasoning scoreBenchLeader55.667.8
Agents & tools scoreBenchLeader–48.9
Multimodal scoreBenchLeader–67.2
Blended price $/MBenchLeader$1.98$1.50
Output speedBenchLeader80 tok/s95 tok/s
Time to first answerBenchLeader42.8 s2.2 s
Context windowBenchLeader1,048,576 tokens1,048,576 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub90.9%94.8%
Arena Hard PromptsArena (LMArena)14821510
LiveBench 1.0 ReasoningLiveBench85.8%†87.8%
GPQA Diamond (AA)Artificial Analysis92.8%†94.5%
Humanity's Last Exam (AA)Artificial Analysis41.5%47.9%
HLE-Diamond (AA)Artificial Analysis20.5%–
GPQA Diamond (Vals)Vals AI92.4%†93.9%
Kagi LLM BenchmarkKagi LLM Benchmark53.5%†–
ARC-AGI-1ARC Prize87.2%95.5%
ARC-AGI-2ARC Prize59.7%84.6%
CritPtArtificial Analysis18.6%14.3%
CritPt v2 hard subset (AA)Artificial Analysis28.5%–
HLE-Diamond (Scale)Scale AI SEAL13.4%†–
Arena ExpertArena (LMArena)14811519
Chess PuzzlesEpoch AI Benchmarking Hub16.0%66.0%
Mystery Game PuzzlesEpoch AI Benchmarking Hub43.0%†37.0%
CL-bench LifeEpoch AI Benchmarking Hub13.5%–
LMCAEpoch AI Benchmarking Hub45.5%†50.4%
DTBenchEpoch AI Benchmarking Hub93.9%†96.8%
ForecastBenchForecastBench60.7%†66.2%
Coding
SWE-bench Verified (Epoch)Epoch AI Benchmarking Hub77.6%†–
SciCodeSciCode46.4%57.2%
WeirdMLWeirdML46.5%–
FrontierCodeCognition28.6%43.6%†
Arena CodingArena (LMArena)15031520
Show all 140 benchmarks
MetricPublisherDeepSeek V4 Pro (high)Gemini 3.7 Flash (high)
Coding
Arena WebDevArena (LMArena)15811588
LiveBench 1.0 CodingLiveBench77.2%†78.9%
SciCode (AA)Artificial Analysis51.7%57.2%
LiveCodeBenchVals AI87.5%†88.7%
IOIVals AI51.6%†67.8%
SWE-bench (Vals)Vals AI96.4%†80.8%
Code MigrationVals AI41.5%†34.8%
Code Migration v1.1Vals AI52.0%†–
ProgramBenchVals AI0.0%†0.0%
Vibe Code Bench 1-100Vals AI17.5%†–
Vibe Code Bench v1.1Vals AI82.3%†70.4%
SWE Atlas v1.0: Codebase QnAScale AI SEAL27.1%†–
SWE Atlas v1.0: Test WritingScale AI SEAL27.1%†–
MirrorCodeEpoch AI Benchmarking Hub33.3%†41.1%
Surface Evolver BenchEpoch AI Benchmarking Hub40.0%–
DeepSWE v1.1Epoch AI Benchmarking Hub–65.3%
ALE-BenchEpoch AI Benchmarking Hub1006.1904.3
FrontierSWEEpoch AI Benchmarking Hub–20.3%
Terminal-Bench 4.0 (AA)Artificial Analysis6.1%13.6%
Terminal-Bench 2.1 (AA)Artificial Analysis78.7%†85.8%
Startrise FrontendStartrise62.7%†–
GameBench 2 OverallDuelLab GameBench 241.9†–
GameBench 2 by reasoning settingDuelLab GameBench 245.1–
Design Arena WebsiteDesign Arena1227†1295†
FrontierCode 1.1 (AA)Artificial Analysis38.8%†–
FrontierCode 2.0 (AA)Artificial Analysis16.7%–
Agents & tools
Terminal-BenchTerminal-Bench6.7%†12.4%
Remote Labor IndexScale AI / CAIS–5.0%†
APEX-AgentsMercor47.3%†67.8%†
Arena AgentArena (LMArena)–-2.2
LiveBench Agentic CodingLiveBench55.0%†58.3%
Terminal-Bench 2.1 (Vals)Vals AI50.2%†77.5%
HiL-BenchScale AI SEAL38.5%†–
GDPval-AA v2.1, normalised scoreArtificial Analysis47.5%†–
GDPval-AA v2.1Artificial Analysis13361404
τ³-Banking (AA)Artificial Analysis39.6%†32.8%
Analyst Agent (AA)Artificial Analysis–60.0%
CyberBenchVals AI72.9%†43.8%
Finance Agent v2Vals AI50.4%†59.0%
Finance Agent v2.1Vals AI57.1%†–
Harvey's Legal Agent Benchmark v1.1Vals AI10.8%†–
Harvey's Legal Agent Benchmark v1Vals AI7.5%†8.8%
Legal Research BenchVals AI40.9%†34.6%
SkillsBenchVals AI53.8%†65.9%
SREBenchVals AI1.9%†4.6%
Tax Agent BenchVals AI26.8%†22.8%
Terminal-Bench 2.0 (Vals)Vals AI56.2%†–
Terminal-Bench 4.0 (Vals)Vals AI14.1%†12.1%
Terminal-Bench ScienceVals AI4.3%†7.1%
Arena Agent Task CompletionArena (LMArena)–-3.5
Vending-Bench 2Andon Labs3284.5†–
AutomationBenchArtificial Analysis46.2%62.0%
GDP.pdf (AA)Artificial Analysis10.4%23.6%
EnterpriseOps-GymArtificial Analysis49.6%†50.4%†
AA-Briefcase v1.1Artificial Analysis11481115
Design Arena AgenticDesign Arena927†1222†
Terminal-Bench-Science 0.1 (AA)Artificial Analysis1.9%–
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub64.6%†71.6%
FrontierMath Tier 4Epoch AI Benchmarking Hub26.8%†36.6%
OTIS Mock AIMEEpoch AI Benchmarking Hub95.6%97.2%
ProofBenchVals AI50.0%†58.0%†
LiveBench MathematicsLiveBench95.1%†93.5%
AIME 2026MathArena96.7%†–
HMMT February 2026MathArena93.9%†–
USAMO 2026MathArena60.7%†–
MathArena ApexMathArena28.1%†–
Arena MathsArena (LMArena)14651502
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub52.9%†69.2%
LiveBench 1.0 Data AnalysisLiveBench79.2%†68.0%
AA-Omniscience v1.1Artificial Analysis-6.5–
MMLU-ProVals AI87.3%†90.1%
LegalBenchVals AI82.4%†87.3%
CorpFinVals AI65.4%†–
TaxEvalVals AI73.1%†74.7%
CaseLaw v2Vals AI59.4%†–
Excel Modeling BenchmarkVals AI52.8%†71.3%
Excel Modeling Benchmark v1.1Vals AI65.1%†–
MedCodeVals AI42.5%†53.4%
MedScribeVals AI80.2%†83.9%
Public Benefits Bench v1.1Vals AI62.9%†–
Immigration BenchVals AI61.7%†–
DrugDiscoveryBenchScale AI SEAL31.7%†–
GDP.pdfEpoch AI Benchmarking Hub–23.8%
AA-Omniscience v1.1: accuracyArtificial Analysis45.1%–
AA-Omniscience v1.1: non-hallucinationArtificial Analysis5.9%–
Instruction following
LiveBench LanguageLiveBench82.1%†85.5%
LiveBench Instruction FollowingLiveBench67.7%†79.9%
Arena Instruction FollowingArena (LMArena)14611486
Human preference
Arena TextArena (LMArena)14641489
EQ-Bench 4EQ-Bench1156†–
EQ-Bench Creative Writing v3EQ-Bench1553†1720†
Arena Creative WritingArena (LMArena)14441487
Arena Multi-turnArena (LMArena)14691490
Arena Longer QueriesArena (LMArena)14741501
Multimodal
Arena VisionArena (LMArena)–1313
MMMU-ProArtificial Analysis–85.5%
MMMU-Pro (Vals)Vals AI–89.0%
MortgageTaxVals AI–66.7%
Humanity's Sixth SenseScale AI SEAL–39.8%
Furniture AssemblyEpoch AI Benchmarking Hub–26.7%
Long context
AA-LCR v1.1Artificial Analysis79.3%81.7%
Arena DocumentArena (LMArena)–1468
MLCRArtificial Analysis–12.2%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub155.9†158.4†
LiveBench 1.0LiveBench77.4%†78.8%
AA Intelligence Index v5.0.1Artificial Analysis20.8–
AA Intelligence Index v4.3.2Artificial Analysis36†39.1
AA Progress Index v1Artificial Analysis311.5†317.4
Vals IndexVals AI53.0†–
Vals Finance IndexVals AI52.9%†–
Vals RSI IndexVals AI–18.3%
Vals Router IndexVals AI53.0%†–
Vals Legal IndexVals AI69.0%†–
Safety & honesty
DistressBenchScale AI SEAL–69.0%†
Metadata
AA Openness IndexArtificial Analysis44.4–

Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.

DuelLab GameBench 2, release gb2-20261111-7b598bb13de4 (https://benchmarks.duellab.org/releases/gb2-20261111-7b598bb13de4/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

What changed

  • Gemini 3.7 Flash (high): 2 new results (GDPval-AA v2.1, AA-Briefcase v1.1).
  • DeepSeek V4 Pro (high): BenchLeader Index 59.1 → 56.0.
  • DeepSeek V4 Pro (high): 17 new results (AA Intelligence Index v5.0.1, Humanity's Last Exam (AA), HLE-Diamond (AA)…); input price moved from $0.66/M to $1.32/M (100% up); output price moved from $1.98/M to $3.96/M (100% up).
  • Gemini 3.7 Flash (high): BenchLeader Index 62.6 → 62.1.
  • DeepSeek V4 Pro (high): BenchLeader Index 57.5 → 59.1.
  • Gemini 3.7 Flash (high): 1 score revision.
  • Gemini 3.7 Flash (high): 1 new result (Arena Document).
  • DeepSeek V4 Pro (high): 1 score revision.
  • DeepSeek V4 Pro (high): 1 score revision.
  • Gemini 3.7 Flash (high): 1 new result (MirrorCode).

Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs

DeepSeek V4 Pro vs Gemini 3.7 Flash: questions

Is Gemini 3.7 Flash better than DeepSeek V4 Pro?
Gemini 3.7 Flash (high) leads on quality: 62.1 vs 56.0. The BenchLeader Index combines every independent quality benchmark; Gemini 3.7 Flash (high) is ahead overall as of 16 Nov 2026, but check the category scores for your use.
DeepSeek V4 Pro vs Gemini 3.7 Flash: which is cheaper, and will that change?
DeepSeek V4 Pro is cheaper today: $0.990 against $1.50 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is DeepSeek V4 Pro or Gemini 3.7 Flash better for coding?
Gemini 3.7 Flash scores higher in coding (62 vs 52 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1581 for DeepSeek V4 Pro against 1588 for Gemini 3.7 Flash.
Which is faster, DeepSeek V4 Pro or Gemini 3.7 Flash?
Gemini 3.7 Flash streams ~2× faster (95 vs 80 output tokens per second, Artificial Analysis 7-day median).
Which is better for long documents?
Gemini 3.7 Flash scores higher on long-context benchmarks (66 vs 64). Both accept 1,048,576 tokens of context.
More questions (3)
Which has lower latency?
Gemini 3.7 Flash answers first: 2.2 s against 42.8 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
Do independent benchmarks agree on DeepSeek V4 Pro vs Gemini 3.7 Flash?
Across the 25 benchmarks both have, DeepSeek V4 Pro leads 3, Gemini 3.7 Flash leads 22. Every overall index that rates both puts Gemini 3.7 Flash ahead.
When was DeepSeek V4 Pro released?
DeepSeek V4 Pro was released on 13 Aug 2026.