priorsBuilder

DeepSeek · released 2026-04-24

DeepSeek V4 Pro

Measured on 18 benchmarks across 3 reasoning efforts.

General frontier

36.5

#68 of 135 models · best at max

includes estimates: 47% of the mix is measured for this model

Reasoning efforts in your mix

max36.5#174 of 312
none31.7#211 of 312
xhigh31.7#212 of 312

Your mix’s benchmarks

4 of 8 measured

reasoning

BullshitBench v2

196 results
  1. #158xhigh of 19614.0%
  2. #162none of 19614.0%

Leader: Opus 4.8 · 95.0%

coding

DeepSWE v1.1

70 results
  1. #31max of 7062.8%

Leader: GPT-6 Astra · 74.1%

math

FrontierMath Tier 4 (v2)

70 results
  1. #64max of 702.4%

Leader: GPT-6.1 Sol · 100.0%

reasoning

SimpleBench

97 results
  1. #52max of 9750.9%

Leader: Opus 5.5 · 88.4%

All 18 benchmarks

show 14 more

agentic

Agent Arena

50 results
  1. #27high of 501.3%

Leader: Opus 5.5 · 15.5%

agentic

Vending-Bench 2

67 results
  1. #47max of 67$3,285

Leader: GPT-6 Astra · $15,515

coding

Arena WebDev

122 results
  1. #31max of 1221582

Leader: Opus 5.5 · 1813

coding

DeepSWE v1

29 results
  1. #25max of 297.5%

Leader: GPT-5.5 · 70.0%

coding

FrontierCode 1.1 Extended

132 results
  1. #123none of 13231.0%

Leader: Opus 5.5 · 65.3%

coding

FrontierCode 1.1 Main

132 results
  1. #123none of 13217.6%

Leader: Opus 5.5 · 54.6%

coding

FrontierSWE v1

17 results
  1. #14max of 1724.6%

Leader: Fable 5 · 88.2%

coding

SWE Atlas-QnA

22 results
  1. #16max of 2227.1%

Leader: Opus 5 · 63.2%

coding

WeirdML v2

127 results
  1. #65max of 12748.9%

Leader: GPT-6 Astra Pro · 93.6%

composite

Arena Text

220 results
  1. #50max of 2201465

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #46max of 2201475

Leader: Gemini 4 · 1553

creative

MineBench

72 results
  1. #39max of 721489

Leader: GPT-6 Astra Pro · 2281

creative

VoxelBench (text)

55 results
  1. #27max of 551632

Leader: GPT-6 Astra · 2733

math

FrontierMath Tiers 1–3 (v2)

115 results
  1. #52max of 11545.3%

Leader: GPT-6.1 Sol · 93.7%