BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases4

DeepSeek V4.1 Flash 1017 vs DeepSeek V4 Flash

Full profiles: DeepSeek V4.1 Flash 1017 (DeepSeek) and DeepSeek V4 Flash (DeepSeek): every benchmark, price and rank.

Updated · Data from 9 sources

As of 2 Nov 2026, DeepSeek V4.1 Flash and DeepSeek V4 Flash are neck and neck on quality (59.2 vs 57.3 on the BenchLeader Index, inside the ±5.3 margin), and DeepSeek V4 Flash is 2.0× cheaper at $0.262 vs $0.525 per 1M tokens blended. Choose V4.1 Flash for coding (64 vs 55), where it also streams 3×+ faster; choose V4 Flash for fast first answers (1.0 s vs 13.0 s) and knowledge (55 vs 48) and reasoning (60 vs 58).

Key facts (as of 2 Nov 2026)

  • DeepSeek V4.1 Flash scores 59.2 on the BenchLeader Index and DeepSeek V4 Flash scores 57.3 (BenchLeader, 2 Nov 2026).
  • DeepSeek V4.1 Flash costs $0.525 per 1M tokens blended (list price, 2 Nov 2026).
  • DeepSeek V4 Flash costs $0.262 per 1M tokens blended (list price, 2 Nov 2026).
  • DeepSeek V4.1 Flash streams 3×+ faster than DeepSeek V4 Flash (Artificial Analysis, 7-day median, 2 Nov 2026).
  • Of the 14 benchmarks both models have, DeepSeek V4.1 Flash leads 14 (BenchLeader, 2 Nov 2026).
  • On Arena WebDev, DeepSeek V4.1 Flash 1017 is rated 1665 and DeepSeek V4 Flash 1582 (LMArena, as published by 2 Nov 2026).
DeepSeek V4.1 Flash 1017 benchmarks max DeepSeek DeepSeek V4 Flash benchmarks high DeepSeekAdd a third model

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
CodingDeepSeek V4.1 Flash 101764 vs 55 category score; Arena WebDev 1665 vs 1582High
MathsEither62 vs 62 category scoreEither
ReasoningDeepSeek V4 Flash60 vs 58 category score; ARC-AGI-2 56.0% vs 78.1%Medium
KnowledgeDeepSeek V4 Flash55 vs 48 category score; Excel Modeling Benchmark 57.0% vs 61.8%Medium
Chat quality (human preference)–Only DeepSeek V4 Flash has an Arena rating (62 category score)Not comparable
Lowest costDeepSeek V4 Flash$0.262 vs $0.525 per 1M blended; agentic coding session $0.011 vs $0.023–
Fastest first answerDeepSeek V4 Flash1.0 s (DeepSeek V4 Flash (high)) vs 13.0 s (DeepSeek V4.1 Flash 1017 (max)) 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

  • DeepSeek V4.1 Flash 1017 (max) and DeepSeek V4 Flash (high) are neck and neck on quality (59.2 vs 57.3, within the ±5.3 margin).
  • DeepSeek V4.1 Flash 1017 (max) is stronger in agents & tools, coding, long context, multimodal.
  • DeepSeek V4 Flash (high) is stronger in knowledge, maths, reasoning, human preference.
  • DeepSeek V4 Flash (high) is 2.0× cheaper ($0.262 vs $0.525 per 1M blended).
  • DeepSeek V4.1 Flash 1017 (max) streams 3.5× faster (170 vs 49 tokens per second).

At a glance

SpecDeepSeek V4.1 Flash 1017DeepSeek V4 Flash
DeveloperDeepSeekDeepSeek
Released17 Oct 202631 Jul 2026
Context window (tokens)1,048,5761,048,576
Compared atDeepSeek V4.1 Flash 1017 (max)DeepSeek V4 Flash (high)
Open weightsYesYes
Price in / out / cached input per 1M$0.300 / $1.20 / $0.0060$0.150 / $0.600 / $0.0030

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 V4.1 Flash 1017 (max)DeepSeek V4 Flash (high)Leader
Reasoning58.159.7DeepSeek V4 Flash
Coding63.755.3DeepSeek V4.1 Flash 1017
Agents & tools63.0–not comparable
Maths61.761.8level
Knowledge48.354.7DeepSeek V4 Flash
Human preference–62.0not comparable
Multimodal63.2–not comparable
Long context66.2–not comparable

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 avg75ReasoningCodingMathsKnowledgeDeepSeek V4.1 Flash 1017, Reasoning: 58.1DeepSeek V4.1 Flash 1017, Coding: 63.7DeepSeek V4.1 Flash 1017, Maths: 61.7DeepSeek V4.1 Flash 1017, Knowledge: 48.3DeepSeek V4 Flash, Reasoning: 59.7DeepSeek V4 Flash, Coding: 55.3DeepSeek V4 Flash, Maths: 61.8DeepSeek V4 Flash, Knowledge: 54.7

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

ModelIndex

Where sources disagree

Across the 14 benchmarks both have, DeepSeek V4.1 Flash 1017 leads 14, DeepSeek V4 Flash leads 0.

Overall indexDeepSeek V4.1 Flash 1017 (max)DeepSeek V4 Flash (high)Leader
BenchLeader Index59.257.3neck and neck

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 V4.1 Flash 1017 (max)DeepSeek V4 Flash (high)
Blended price per 1M tokens$0.525$0.262
Input / output per 1M$0.300 / $1.20$0.150 / $0.600
Cached input per 1M$0.0060$0.0030
Price per task: chat reply$0.0005$0.0002
Price per task: summarize a 30-page report$0.0043$0.0022
Price per task: code edit$0.0036$0.0018
Price per task: agentic coding session$0.023$0.011
Price per task: structured extraction$0.0008$0.0004
Agentic coding session, input mostly cached$0.0096$0.0048
Cost per benchmark task (AA Index)$0.602–
Output tokens per task (AA Index)139k–
Of which reasoning91k–
Time per task (AA Index)10 min–
Output speed170 tok/s49 tok/s
Time to first answer13.0 s1.0 s
Time to a complete answer15.9 s–

Each model is compared at its best-ranked reasoning-effort setting. DeepSeek V4.1 Flash 1017 (max) scores 59.2 and takes 13 s to its first answer token; DeepSeek V4.1 Flash 1017 (high) scores 51.5 in 1 s, DeepSeek V4.1 Flash 1017 (no reasoning) scores 45.0 in 1 s. DeepSeek V4 Flash (high) scores 57.3 and takes 1 s to its first answer token; DeepSeek V4 Flash (max) scores 57.0 in 15 s, DeepSeek V4 Flash scores 54.3 in 1 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

122 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 V4.1 Flash 1017 (max)DeepSeek V4 Flash (high)
BenchLeader IndexBenchLeader59.257.3
Agents & tools scoreBenchLeader63.0–
Coding scoreBenchLeader63.755.3
Knowledge scoreBenchLeader48.354.7
Long context scoreBenchLeader66.2–
Maths scoreBenchLeader61.761.8
Multimodal scoreBenchLeader63.2–
Reasoning scoreBenchLeader58.159.7
Human preference scoreBenchLeader–62.0
Blended price $/MBenchLeader$0.525$0.262
Output speedBenchLeader170 tok/s49 tok/s
Time to first answerBenchLeader13.0 s1.0 s
Context windowBenchLeader1,048,576 tokens1,048,576 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub93.2%91.0%†
SimpleBenchSimpleBench–61.1%†
Arena Hard PromptsArena (LMArena)–1458
LiveBench 2.0 ReasoningLiveBench89.1%†–
LiveBench 1.0 ReasoningLiveBench–86.6%†
GPQA Diamond (AA)Artificial Analysis–90.8%†
Humanity's Last Exam (AA)Artificial Analysis43.3%38.5%†
HLE-Diamond (AA)Artificial Analysis25.3%–
GPQA Diamond (Vals)Vals AI–89.9%
Kagi LLM BenchmarkKagi LLM Benchmark–52.2%†
ARC-AGI-1ARC Prize93.5%87.0%
ARC-AGI-2ARC Prize78.1%56.0%
CritPtArtificial Analysis24.0%16.6%†
CritPt v2 hard subset (AA)Artificial Analysis24.0%–
Chess PuzzlesEpoch AI Benchmarking Hub46.0%32.0%†
Mystery Game PuzzlesEpoch AI Benchmarking Hub49.0%34.0%†
LMCAEpoch AI Benchmarking Hub–41.7%†
DTBenchEpoch AI Benchmarking Hub–90.9%†
Coding
SciCodeSciCode55.9%42.0%
WeirdMLWeirdML–57.0%
FrontierCodeCognition–18.8%†
Arena CodingArena (LMArena)–1480
Arena WebDevArena (LMArena)16651582
LiveBench 2.0 CodingLiveBench96.0%†–
LiveBench 1.0 CodingLiveBench–75.0%†
Show all 122 benchmarks
MetricPublisherDeepSeek V4.1 Flash 1017 (max)DeepSeek V4 Flash (high)
Coding
SciCode (AA)Artificial Analysis55.9%50.4%†
LiveCodeBenchVals AI–87.3%
IOIVals AI–32.7%
SWE-bench (Vals)Vals AI–88.8%
Code MigrationVals AI44.2%38.6%
Code Migration v1.1Vals AI67.5%–
ProgramBenchVals AI–0.0%
Vibe Code Bench v1.1Vals AI87.7%74.7%
CursorBenchCursor47.7%–
ALE-BenchEpoch AI Benchmarking Hub1622678.2
Terminal-Bench 4.0 (AA)Artificial Analysis33.8%12.1%†
Terminal-Bench 2.1 (AA)Artificial Analysis–78.7%†
Startrise FrontendStartrise74.2%†–
GameBench 2 OverallDuelLab GameBench 263.4†34.2†
GameBench 2 by reasoning settingDuelLab GameBench 271.944.0
FrontierCode 1.1 (AA)Artificial Analysis43.7%–
FrontierCode 2.0 (AA)Artificial Analysis28.5%–
Agents & tools
APEX-AgentsMercor69.8%–
Arena AgentArena (LMArena)2.7–
LiveBench Long AgenticLiveBench35.7%†–
LiveBench BrowserLiveBench67.2%†–
LiveBench Agentic CodingLiveBench–46.8%†
Terminal-Bench 2.1 (Vals)Vals AI–67.0%†
HiL-BenchScale AI SEAL–47.8%†
GDPval-AA v2.1, normalised scoreArtificial Analysis62.3%–
GDPval-AA v2.1Artificial Analysis1746–
τ³-Banking (AA)Artificial Analysis–39.4%†
CyberBenchVals AI–72.1%
Finance Agent v2Vals AI57.0%49.5%
Finance Agent v2.1Vals AI62.9%–
Harvey's Legal Agent Benchmark v1.1Vals AI14.2%–
Harvey's Legal Agent Benchmark v1Vals AI9.2%8.3%
Legal Research BenchVals AI45.2%30.3%
SkillsBenchVals AI–50.7%
Tax Agent BenchVals AI44.4%28.1%
Terminal-Bench 4.0 (Vals)Vals AI31.8%18.7%
Terminal-Bench ScienceVals AI–4.3%
AutomationBenchArtificial Analysis66.6%54.0%†
GDP.pdf (AA)Artificial Analysis17.2%11.0%†
AA-Briefcase v1.1Artificial Analysis1570–
GDPxGDPx by SemiAnalysis5.4%–
Terminal-Bench-Science 0.1 (AA)Artificial Analysis12.9%–
CWE-Bench (AA)Artificial Analysis69.2%–
DeepSec-Bench (AA)Artificial Analysis24.9%–
CyberGym end-to-end (AA)Artificial Analysis76.3%–
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub71.2%57.5%†
FrontierMath Tier 4Epoch AI Benchmarking Hub46.3%24.4%†
OTIS Mock AIMEEpoch AI Benchmarking Hub99.4%94.4%†
ProofBenchVals AI–56.0%†
LiveBench MathematicsLiveBench–86.8%†
AIME 2026MathArena–95.8%†
HMMT February 2026MathArena–93.9%†
MathArena ApexMathArena–27.1%†
Arena MathsArena (LMArena)–1442
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub42.5%33.6%†
LiveBench 2.0 Data AnalysisLiveBench79.5%†–
LiveBench ResearchLiveBench79.5%†–
LiveBench 1.0 Data AnalysisLiveBench–79.3%†
AA-Omniscience v1.1Artificial Analysis0.9–
MMLU-ProVals AI–86.2%
LegalBenchVals AI–77.7%
CorpFinVals AI–61.9%
TaxEvalVals AI–70.7%
PRBench FinanceScale AI SEAL47.1%†–
PRBench LegalScale AI SEAL47.3%†–
BioMysteryBenchVals AI–64.4%
Excel Modeling BenchmarkVals AI61.8%57.0%
Excel Modeling Benchmark v1.1Vals AI69.2%–
MedCodeVals AI–41.4%
MedScribeVals AI–80.4%
DrugDiscoveryBenchScale AI SEAL60.6%†–
AA-Omniscience v1.1: accuracyArtificial Analysis39.6%–
AA-Omniscience v1.1: non-hallucinationArtificial Analysis35.9%–
Instruction following
LiveBench LanguageLiveBench–79.2%†
LiveBench Instruction FollowingLiveBench–65.5%†
MultiChallengeScale AI SEAL74.9%†–
Arena Instruction FollowingArena (LMArena)–1434
Human preference
Arena TextArena (LMArena)–1439
Arena Creative WritingArena (LMArena)–1410
Arena Multi-turnArena (LMArena)–1447
Arena Longer QueriesArena (LMArena)–1448
Multimodal
LiveBench DesignLiveBench83.4%†–
MMMU-ProArtificial Analysis80.6%–
Long context
AA-LCR v1.1Artificial Analysis82.7%79.7%†
MLCRArtificial Analysis26.7%–
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub158.8†154.9†
LiveBench 2.0LiveBench72.9%†–
LiveBench 1.0LiveBench–74.2%†
AA Intelligence Index v5.0Artificial Analysis32.5–
AA Intelligence Index v4.3.2Artificial Analysis45.134.3†
AA Progress Index v1Artificial Analysis323.7309.4†
Vals IndexVals AI61.3–
Vals Finance IndexVals AI60.9%–
Vals Router IndexVals AI61.3%†–
AA Cyber IndexArtificial Analysis56.8–
Safety & honesty
FORTRESSScale AI SEAL27.2%†–
Metadata
AA Openness IndexArtificial Analysis44.4–

Data as of 2026-11-02. Best configuration of each model; every score links to its source on the model pages.

DuelLab GameBench 2, release gb2-20261026-cb69502a12d9 (https://benchmarks.duellab.org/releases/gb2-20261026-cb69502a12d9/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

What changed

  • DeepSeek V4 Flash (high): 1 new result (GameBench 2 by reasoning setting).
  • DeepSeek V4.1 Flash 1017 (max): 1 score revision.
  • DeepSeek V4.1 Flash 1017 (max): 1 score revision.
  • DeepSeek V4.1 Flash 1017 (max): 1 new result (Arena Agent); 1 score revision.
  • DeepSeek V4.1 Flash 1017 (max): 4 new results (SciCode, APEX-Agents, CursorBench…); 1 score revision.
  • DeepSeek V4.1 Flash 1017 (max): BenchLeader Index 61.8 → 59.0.
  • DeepSeek V4.1 Flash 1017 (max): 5 new results (AA Intelligence Index v5.0, AA Progress Index v1, HLE-Diamond (AA)…); 4 score revisions; cost per task moved from $0.26/M to $0.60/M (130% up).
  • DeepSeek V4.1 Flash 1017 (max): BenchLeader Index 61.3 → 61.8.
  • DeepSeek V4.1 Flash 1017 (max): 10 new results (MLCR, AA Cyber Index, CWE-Bench (AA)…).
  • DeepSeek V4 Flash (high): 1 new result (Harvey's Legal Agent Benchmark v1).

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

DeepSeek V4.1 Flash 1017 vs DeepSeek V4 Flash: questions

Is DeepSeek V4.1 Flash 1017 better than DeepSeek V4 Flash?
DeepSeek V4.1 Flash 1017 (max) and DeepSeek V4 Flash (high) are neck and neck on quality (59.2 vs 57.3, within the ±5.3 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 2 Nov 2026, so check the category scores for your use.
DeepSeek V4.1 Flash 1017 vs DeepSeek V4 Flash: which is cheaper, and will that change?
DeepSeek V4 Flash is cheaper today: $0.262 against $0.525 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is DeepSeek V4.1 Flash 1017 or DeepSeek V4 Flash better for coding?
DeepSeek V4.1 Flash 1017 scores higher in coding (64 vs 55 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1665 for DeepSeek V4.1 Flash 1017 against 1582 for DeepSeek V4 Flash.
Which is faster, DeepSeek V4.1 Flash 1017 or DeepSeek V4 Flash?
DeepSeek V4.1 Flash 1017 streams 3×+ faster (170 vs 49 output tokens per second, Artificial Analysis 7-day median).
Is DeepSeek V4.1 Flash the same as DeepSeek V4.1 Flash 1017?
DeepSeek V4.1 Flash 1017 is the checkpoint compared here (BenchLeader Index 59.2); people searching for DeepSeek V4.1 Flash land on this page because it replaced that earlier build.
More questions (4)
Which is better for long documents?
Both accept 1,048,576 tokens of context.
Which has lower latency?
DeepSeek V4 Flash answers first: 1.0 s against 13.0 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 V4.1 Flash 1017 vs DeepSeek V4 Flash?
Across the 14 benchmarks both have, DeepSeek V4.1 Flash 1017 leads 14, DeepSeek V4 Flash leads 0.
When was DeepSeek V4.1 Flash 1017 released?
DeepSeek V4.1 Flash 1017 was released on 17 Oct 2026.