supermarket

Language

DeepSeek-V4-Pro

By DeepSeek, in the DeepSeek-V family.

Released
24 Apr 2026
Status
Current
Context
1M
Max output
393.2K
Input per 1M tokens
$1.32
Output per 1M tokens
$3.96
Open weights
Yes
License
mit
Parameters
1600B, 49B active
Knowledge cutoff
—
Reasoning
Yes
Modalities
text → text

Category scores

Out of 100, from independent and third-party results only, by method category-2026-09-29.1.

Category scores
CategoryScoreRankCoverage
Knowledge & reasoning48.9192 of 4
Coding & software engineering28.6Not ranked1 of 5
Factuality & hallucination52.9181 of 1
Human preference66.6Not ranked2 of 5

Benchmark results

Every result we hold for this model, with who ran it and where it was published.

Knowledge & reasoning

Knowledge & reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Chess PuzzlesPuzzles solved47.0%Independentmax18 Aug 2026Epoch AI
Chess PuzzlesPuzzles solved20.0%Independentmax16 Jun 2026Epoch AI
Chess PuzzlesPuzzles solved13.0%Independenthigh6 Aug 2026Epoch AI
Chess PuzzlesPuzzles solved7.0%Independentnone6 Aug 2026Epoch AI
DTBench (Decision Theory Benchmark)Accuracy (capability questions)93.9%Third partyunknown—https://conceptualreasoning.ai/dtbench via Epoch AI
DTBench (Decision Theory Benchmark)Accuracy (capability questions)90.7%Third partymax—https://conceptualreasoning.ai/dtbench via Epoch AI
LMCA (Language Model Conceptual Argumentation)Correlation with expert ratings (Pearson r)0.4545Third partyunknown—https://conceptualreasoning.ai/lmca via Epoch AI
LMCA (Language Model Conceptual Argumentation)Correlation with expert ratings (Pearson r)0.412Third partymax—https://conceptualreasoning.ai/lmca via Epoch AI
Mystery Game PuzzlesPuzzles solved43.0%Independentmax19 Aug 2026Epoch AI
Mystery Game PuzzlesPuzzles solved17.0%Independentnone27 Aug 2026Epoch AI

Scientific reasoning

Scientific reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
GPQA DiamondAccuracy91.7%Independentmax18 Aug 2026Epoch AI
GPQA DiamondAccuracy90.9%Independenthigh6 Aug 2026Epoch AI
GPQA DiamondAccuracy89.6%Independentmax16 Jun 2026Epoch AI
GPQA DiamondAccuracy73.2%Independentnone6 Aug 2026Epoch AI
Surface Evolver BenchMean score (partial credit)40.0%Third partyhigh—https://yhenon.github.io/surface-evolver-llm-eval/ via Epoch AI

Mathematics

Mathematics results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
FrontierMath Tier 4Accuracy26.8%Independentmax19 Aug 2026Epoch AI
FrontierMath Tier 4Accuracy2.4%Independentmax17 Jun 2026Epoch AI
FrontierMath Tiers 1-3Accuracy64.6%Independentmax19 Aug 2026Epoch AI
FrontierMath Tiers 1-3Accuracy45.3%Independentmax17 Jun 2026Epoch AI
OTIS Mock AIME 2024-2025Accuracy98.6%Independentmax18 Aug 2026Epoch AI
OTIS Mock AIME 2024-2025Accuracy96.7%Independentmax17 Jun 2026Epoch AI
OTIS Mock AIME 2024-2025Accuracy95.6%Independenthigh6 Aug 2026Epoch AI
OTIS Mock AIME 2024-2025Accuracy46.7%Independentnone6 Aug 2026Epoch AI

Coding & software engineering

Coding & software engineering results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
FrontierCodeMain score (Mean@5)28.6%Third partyhigh—https://cognition.com/frontiercode via Epoch AI
FrontierCodeMain score (Mean@5)17.6%Third partynone—https://cognition.com/frontiercode via Epoch AI
SWE-bench Verified% resolved77.6%Independentmax18 Jun 2026Epoch AI
WeirdML (v2)Average accuracy66.2%Third partymax—https://htihle.github.io/weirdml.html via Epoch AI
WeirdML (v2)Average accuracy48.9%Third partymax—https://htihle.github.io/weirdml.html via Epoch AI
WeirdML (v2)Average accuracy46.5%Third partyhigh—https://htihle.github.io/weirdml.html via Epoch AI

Professional knowledge work

Professional knowledge work results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
APEX-AgentsPass@1 (all rubric criteria met)47.3%Source not statedunknown—Epoch AI via Epoch AI

Long context

Long context results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
CL-bench LifeTasks solved13.5%Source not statedhigh—Epoch AI via Epoch AI

Factuality & hallucination

Factuality & hallucination results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
SimpleQA VerifiedAccuracy (% correct)52.9%Independentmax27 Aug 2026Epoch AI
SimpleQA VerifiedAccuracy (% correct)47.0%Independentmax27 Aug 2026Epoch AI

Human preference

Human preference results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Code Arena: WebDev (arena.ai)Arena score (Elo scale)1582Independent—25 Sep 2026Arena
Code Arena: WebDev (arena.ai)Arena score (Elo scale)1446Independent—25 Sep 2026Arena
Code Arena: WebDev (arena.ai)Votes14064Independent—25 Sep 2026Arena
Code Arena: WebDev (arena.ai)Votes4897Independent—25 Sep 2026Arena
Text Arena (arena.ai)Arena score (Elo scale)1464Independent—25 Sep 2026Arena
Text Arena (arena.ai)Arena score (Elo scale)1458Independent—25 Sep 2026Arena
Text Arena (arena.ai)Votes57564Independent—25 Sep 2026Arena
Text Arena (arena.ai)Votes9857Independent—25 Sep 2026Arena

Other DeepSeek-V versions

Sources