priorsBuilder

DeepSeek · released 2026-09-10

DeepSeek V4.1 Flash

Measured on 71 benchmarks across 3 reasoning efforts.

General frontier

66.5

#7 of 32 models · best at low

Reasoning efforts in your mix

low66.5#20 of 128
max38.1#94 of 128
none17.6#126 of 128

Your mix’s benchmarks

5 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #89max of 58739.2%
  2. #301none of 58710.8%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #160max of 526-5.3
  2. #187none of 526-10.2

Leader: Opus 5.5 · 46.4

reasoning

BullshitBench v2

196 results
  1. #58max of 19657.0%
  2. #59low of 19657.0%

Leader: Opus 4.8 · 95.0%

math

FrontierMath Tier 4 (v2)

70 results
  1. #38max of 7026.8%

Leader: GPT-6.1 Sol · 100.0%

reasoning

SimpleBench

97 results
  1. #23max of 9766.7%

Leader: Opus 5.5 · 88.4%

All 71 benchmarks

show 66 more

agentic

AA-Briefcase v1.1

224 results
  1. #53max of 2241422
  2. #97none of 2241118

Leader: Sonnet 5.5 · 1823

agentic

Agent Arena

50 results
  1. #17max of 503.8%

Leader: Opus 5.5 · 15.5%

agentic

AutomationBench-AA

219 results
  1. #5max of 21968.9%
  2. #80none of 21947.8%

Leader: Gemini 4 · 77.5%

agentic

BioMysteryBench

27 results
  1. #16high of 2767.8%

Leader: Sonnet 5.5 · 81.1%

agentic

Harvey's Legal Agent Benchmark

79 results
  1. #29max of 796.7%

Leader: Muse Spark 1.2 · 25.4%

agentic

ITBench-AA

45 results
  1. #10max of 4546.9%

Leader: GPT-5.6 Sol · 56.2%

agentic

Legal Research Bench

78 results
  1. #25max of 7841.3%

Leader: Muse Spark 1.3 · 55.3%

agentic

SkillsBench

34 results
  1. #1high of 3469.8%

Leader: this model · 69.8%

agentic

Tax Agent Bench

69 results
  1. #33max of 6926.6%

Leader: Fable 5.1 · 49.2%

agentic

Terminal-Bench 4.0 (Vals)

47 results
  1. #25high of 4719.7%

Leader: Opus 5.5 · 65.2%

agentic

Terminal-Bench Science (Vals)

40 results
  1. #39high of 400.0%

Leader: GPT-6 Astra · 62.9%

agentic

WeirdML v3

17 results
  1. #16high of 176.2%

Leader: GPT-6 Astra · 42.2%

coding

Arena Image-to-WebDev

58 results
  1. #11max of 581609

Leader: Opus 5.5 · 1749

coding

Arena WebDev

122 results
  1. #22max of 1221619

Leader: Opus 5.5 · 1813

coding

Code Migration

77 results
  1. #17high of 7745.6%

Leader: Sonnet 5.5 · 69.8%

coding

IOI

44 results
  1. #35high of 4440.3%

Leader: Gemini 4 · 100.0%

coding

ProgramBench v1 Almost Resolved

62 results
  1. #16max of 6216.5%

Leader: Opus 5.5 · 65.0%

coding

ProgramBench v1 Fully Resolved

62 results
  1. #18max of 620.5%

Leader: Opus 5.5 · 18.5%

coding

ProgramBench v1 Raw Pass Rate

62 results
  1. #15max of 6272.0%

Leader: Opus 5.5 · 87.0%

coding

SciCode

226 results
  1. #86max of 22651.9%
  2. #195none of 22635.5%

Leader: Opus 5.5 · 66.9%

coding

Terminal-Bench 2.1 (Vals)

74 results
  1. #18high of 7474.5%

Leader: Opus 5.5 · 87.6%

coding

Terminal-Bench Science-AA

46 results
  1. #27max of 469.0%

Leader: GPT-6 Astra · 63.3%

coding

Terminal-Bench v4-AA

216 results
  1. #47max of 21626.8%
  2. #97none of 2165.6%

Leader: Sonnet 5.5 · 63.6%

coding

Vibe Code Bench 1-100

22 results
  1. #13high of 2216.4%

Leader: Opus 5.5 · 30.4%

coding

Vibe Code Bench v1.1

105 results
  1. #16high of 10584.7%

Leader: Sonnet 5.5 · 92.4%

composite

AA Intelligence Index v4.3.2

610 results
  1. #66max of 61039.5
  2. #169none of 61024.7

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #38max of 2201475

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #60max of 2201466

Leader: Gemini 4 · 1553

composite

Vals Index

47 results
  1. #26high of 4751.3%

Leader: Gemini 4 · 68.9%

cost

AA Intelligence Index Cost per Task

177 results
  1. #51none of 177$0.15
  2. #66max of 177$0.27

Leader: GPT-6 Luna · $0.00

cost

Cost to Run AA Intelligence Index

177 results
  1. #36none of 177$156
  2. #69max of 177$477

Leader: GPT-6 Luna · $10.63

creative

MineBench

72 results
  1. #12max of 721857

Leader: GPT-6 Astra Pro · 2281

creative

threejseval

34 results
  1. #8max of 341682

Leader: Opus 5.5 · 2079

creative

VoxelBench (text)

55 results
  1. #14max of 551854

Leader: GPT-6 Astra · 2732

economics

Excel Modeling Benchmark

74 results
  1. #39high of 7457.2%

Leader: Fable 5.1 · 76.7%

economics

Finance Agent v2

78 results
  1. #32high of 7853.5%

Leader: Gemini 4 · 65.4%

economics

GDP.pdf

210 results
  1. #105max of 21012.8%
  2. #168none of 2103.8%

Leader: Muse Wonder Full (Max) · 32.6%

economics

GDPval-AA v2.1

288 results
  1. #32max of 2881600
  2. #88none of 2881341

Leader: Opus 5.5 · 1866

knowledge

AA-Omniscience Accuracy

526 results
  1. #79max of 52646.4%
  2. #193none of 52628.4%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #134none of 52646.0%
  2. #517max of 5263.5%

Leader: MiniCPM5-1B (Non-reasoning) · 99.1%

knowledge

Harvey LAB-AA v1.1

27 results
  1. #15max of 271.7%

Leader: Muse Wonder Full (Max) · 13.1%

knowledge

MedCode

94 results
  1. #53high of 9441.2%

Leader: Opus 5 · 63.6%

knowledge

MedScribe

96 results
  1. #25high of 9685.5%

Leader: Opus 5.5 · 91.4%

knowledge

MMMU-Pro

265 results
  1. #81max of 26577.0%
  2. #168none of 26565.6%

Leader: Opus 5.5 · 87.7%

knowledge

Public Benefits Bench v1.1

50 results
  1. #23max of 5064.3%

Leader: Opus 5 · 76.9%

long-context

AA-LCR v1.1

523 results
  1. #14max of 52384.0%
  2. #285none of 52355.3%

Leader: Kimi K3 · 88.7%

long-context

MLCR-AA

94 results
  1. #24max of 9422.8%

Leader: Sonnet 5.5 · 75.0%

math

FrontierMath Tiers 1–3 (v2)

115 results
  1. #28max of 11567.4%

Leader: GPT-6.1 Sol · 93.7%

math

ProofBench v1.1

52 results
  1. #30high of 5254.0%

Leader: Sonnet 5.5 · 100.0%

other

Arena Vision

116 results
  1. #45max of 1161260

Leader: Fable 5 · 1308

reasoning

ARC-AGI-1

215 results
  1. #39max of 21594.5%
  2. #69low of 21590.5%
  3. #78high of 21588.5%
  4. #198none of 2157.8%

Leader: Fable 5 · 98.5%

reasoning

ARC-AGI-2

216 results
  1. #47max of 21672.9%
  2. #59high of 21667.5%
  3. #75low of 21660.6%
  4. #189none of 2161.3%

Leader: GPT-6 Astra · 95.0%

reasoning

CritPt

529 results
  1. #93max of 52914.3%
  2. #264none of 5290.3%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

Furniture Assembly

31 results
  1. #19max of 3134.2%

Leader: Opus 5.5 · 83.3%

reasoning

LegalBench

135 results
  1. #55high of 13583.3%

Leader: Fable 5 · 88.6%

reasoning

MysteryMechanism

27 results
  1. #21high of 2721.2%

Leader: GPT-6 Astra · 53.2%

reasoning

SAGE

80 results
  1. #29high of 8047.9%

Leader: Opus 4.7 · 56.1%

reliability

Tool-call error rate

171 results
  1. #56max of 1710.7%

Leader: GPT-5.4 Pro · 0.0%

reliability

Uptime (7d)

181 results
  1. #115max of 18199.0%

Leader: GPT-5 Pro · 100.0%

safety

CWE-Bench

19 results
  1. #2max of 1968.3%

Leader: Grok 4.7 · 68.3%

safety

CyberBench v1.1

45 results
  1. #10high of 4573.7%

Leader: Wonder · 80.4%

safety

SRE Bench

32 results
  1. #25max of 320.8%

Leader: GPT-6 Astra · 56.9%

speed

Latency (Time to First Token)

351 results
  1. #73max of 3511.0s
  2. #81none of 3511.1s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #48max of 351194.8
  2. #50none of 351193.5

Leader: Celeris-1 · 1522.5

speed

Provider latency

177 results
  1. #46max of 1770.8s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #15max of 176118.8

Leader: Muse Spark 1.1 · 204.3