Benchmark Almanac

DeepSeek-V4-Pro

DeepSeek-V4-Pro ranks first on 5 of the 170 benchmarks it has scores on.

DeepSeek Released 24 Apr 2026 Family DeepSeek-V4 The lab's model card
179
Benchmarks with a score
5
Ranked first, out of five or more models
15
In the top three
DeepSeek Apr 2026

DeepSeek-V4-Pro

Open weights Flagship

149.1 #67 of 271
Epoch Capabilities Index
  • SWE-bench Verified 77.6% #6 /32
  • DeepSWE 62.8% #14 /28
  • Toolathlon 52.8% #5 /50

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

BenchmarkScoreRankSource
AgentIF (2026-10 setting) 67.4% #1 of 24 Official leaderboard
MMLU-Pro 87.5% #7 of 151 Model card
C-Eval 93.1% #3 of 38 Model card
EQ-Bench 3 1,555 #7 of 78 Official leaderboard
MMLU 90.1% #13 of 127 Model card
CMMLU 90.8% #2 of 17 Model card
MMMLU 90.3% #8 of 62 Model card
LMArena (Elo) 1,465 #45 of 333 Official leaderboard
AGIEval 83.1% #5 of 35 Model card
Chinese SimpleQA 84.4% #2 of 13 Model card
BIG-Bench Hard 87.5% #13 of 71 Model card
Extended NYT Connections 91.3% #19 of 97 Official leaderboard
Longform Creative Writing (EQ-Bench) 75.6 #26 of 119 Official leaderboard
DROP 88.7% #9 of 40 Model card
TriviaQA 85.6% #9 of 39 Model card
HellaSwag 88% #11 of 47 Model card
HELM Arabic Enterprise 72.8 #6 of 23 Official leaderboard
Creative Writing v3 75.0 #32 of 120 Official leaderboard
Humanity's Last Exam 48.2% #43 of 147 Model card
MRCR v2 83.5% #8 of 27 Model card
Wer wird Millionär? LLM Benchmark 73489.0 #23 of 76 Official leaderboard
OpenAI MRCR v2 83.5% #3 of 9 Model card
WinoGrande 81.5% #15 of 42 Model card
SEA-HELM 72.6 #24 of 62 Official leaderboard
Vectara Hallucination Leaderboard (2025 dataset) 8.60% #44 of 104 Official leaderboard
SimpleQA Verified 47% #36 of 84 Epoch AI
MGSM 84.4% #26 of 52 Model card
CL-bench Life 13.5% #8 of 16 Official leaderboard
CorpusQA 1M 62% #1 of 2 Model card
MultiLoKo 51.1% #1 of 2 Model card
MMLU-Redux 90.8% #21 of 38 Model card
Continual Learning Exam 27.3% #6 of 10 Official leaderboard
CLUEWSC 85.2% #5 of 8 Model card
Judgemark v4 47.1 #27 of 40 Official leaderboard
EQ-Bench 4 1,156 #22 of 32 Official leaderboard
LongBench v2 51.5% #19 of 27 Model card
BullshitBench V2 14% #100 of 122 Official leaderboard
BullshitBench 14.6% #90 of 107 Official leaderboard
SuperGPQA 53.9% #22 of 26 Model card
Phare 58.9% #53 of 62 Official leaderboard
HLE-Diamond 16.6% #18 of 18 Official leaderboard
DeepSeek Chinese functional writing evaluation 62.6% #1 of 1 Model card
DeepSeek complex instruction following and multi-turn writing evaluation 45.9% #1 of 1 Model card
DeepSeek search Q&A evaluation 28.1% #1 of 1 Model card
FACTS Parametric 62.6% #1 of 1 Model card

Math and science

BenchmarkScoreRankSource
ResearchClawBench 38.6 #1 of 23 Official leaderboard
Echo General AI Prediction Leaderboard 1,018 #1 of 17 Official leaderboard
NatureBench 25.6% #2 of 18 Official leaderboard
IMO-AnswerBench 89.8% #5 of 37 Model card
GPQA Diamond 90.9% #33 of 219 Epoch AI
OTIS Mock AIME 2024-2025 96.7% #36 of 202 Epoch AI
MathArena (Apex, IMO, USAMO) 28.1% #10 of 47 Official leaderboard
DTBench 84.5% #44 of 178 Official leaderboard
ForecastBench 60.6 #25 of 86 Official leaderboard
GSM8K 92.6% #27 of 87 Model card
CritPt 12.9% #51 of 160 Epoch AI
AIME 2026 96.7% #10 of 31 Official leaderboard
CMATH 90.9% #2 of 6 Model card
HMMT 2026 93.9% #12 of 31 Official leaderboard
USAMO 2026 60.7% #4 of 9 Official leaderboard
Putnam 2025 120.0 #1 of 2 Model card
MATH 64.5% #48 of 93 Model card
Verifying the Verifiers (Project APE CRED benchmark) 83.5% #12 of 23 Official leaderboard
SimpleBench 50.9% #53 of 100 Official leaderboard
FrontierMath Tiers 1-3 v2 45.3% #57 of 89 Epoch AI
ProofBench 10% #43 of 59 Official leaderboard
xbench-ScienceQA 69.4% #22 of 29 Official leaderboard
Surface Evolver Bench 40% #20 of 26 Epoch AI
FrontierPhysics 0.60% #16 of 19 Official leaderboard
Leipzig Benchmark 10% #16 of 19 Official leaderboard
ProofBench v1.1 16% #44 of 51 Official leaderboard
Terminal-Bench-Science 3.81% #16 of 17 Official leaderboard
FrontierMath Tier 4 v2 2.44% #67 of 71 Epoch AI
ExplorationBench 12.9% #10 of 10 Official leaderboard

Coding

BenchmarkScoreRankSource
LiveCodeBench v6 93.5% #1 of 76 Model card
LiveCodeBench 93.5% #1 of 73 Model card
Codeforces rating 3,206 #3 of 41 Model card
BigCodeBench 59.2% #2 of 11 Model card
SWE-bench Verified 77.6% #6 of 32 Epoch AI
IOI (Vals) 35.8% #15 of 55 Official leaderboard
Code Arena 1,582 #34 of 124 Official leaderboard
SWE-bench Multilingual 76.2% #26 of 70 Model card
EvoCode-Bench v2 14.1% #5 of 13 Official leaderboard
Monthly-SWEBench 57% #5 of 12 Official leaderboard
WeirdML 48.9% #49 of 112 Official leaderboard
ALE-Bench 1,006 #53 of 121 Epoch AI
EvoCode-Bench 30.6% #6 of 13 Official leaderboard
WebDev Arena 1,464 #51 of 105 Epoch AI
DeepSWE 62.8% #14 of 28 Official leaderboard
Vibe Code Bench 49.9% #59 of 115 Official leaderboard
HumanEval 76.8% #54 of 95 Model card
Code Migration 26.2% #46 of 78 Official leaderboard
Tencent WorkBuddy Bench 75.6% #6 of 10 Official leaderboard
SWE-Bench Pro 55.4% #47 of 77 Model card
SWE Atlas 27.1% #16 of 22 Official leaderboard
AndroidBench 59.5% #24 of 32 Official leaderboard
SWE-Marathon 4.00% #9 of 12 Official leaderboard
SWE-rebench 40.2% #10 of 13 Official leaderboard
FrontierCode 17.6% #40 of 46 Epoch AI
OpenHands Index 40.7% #29 of 33 Official leaderboard
RSI-Exam 32.3% #16 of 18 Official leaderboard
SWE Atlas Test Writing 27.1% #20 of 22 Official leaderboard
SWE-Marathon v1.1 10.6% #17 of 17 Official leaderboard
MLS-Bench 24.4% #15 of 15 Official leaderboard
DeepSeek R&D Coding Benchmark 67% #1 of 1 Model card

Agents and tools

BenchmarkScoreRankSource
Toolathlon 52.8% #5 of 50 Official leaderboard
PACT 1,570 #3 of 26 Official leaderboard
VitaBench 2.0 47.2% #3 of 20 Official leaderboard
Terminal-Bench 2.0 67.9% #11 of 61 Model card
GAIA 74.8% #13 of 57 Official leaderboard
LisanBench 3256.3 #35 of 133 Official leaderboard
BrowseComp 83.4% #22 of 77 Model card
Long-Horizon-Terminal-Bench 2.0 52.9% #6 of 15 Official leaderboard
Claw-Eval 77.2% #10 of 23 Official leaderboard
Claw-Eval Pass^3 59.8% #10 of 23 Official leaderboard
SkillsBench 50.1% #11 of 21 Official leaderboard
DeepTerminalBench 32% #5 of 9 Official leaderboard
ClawBench V2 33.9% #4 of 7 Official leaderboard
MCP Atlas 73.6% #23 of 40 Model card
Agent Arena -0.00 #31 of 53 Official leaderboard
Long-Horizon-Terminal-Bench (LHTB) 30.7% #17 of 25 Official leaderboard
Toolathlon-Verified 55.9% #17 of 25 Official leaderboard
LiveClawBench 65.6% #12 of 17 Official leaderboard
Mystery Game Puzzles 17% #57 of 80 Epoch AI
PinchBench 2.0 61.1% #41 of 57 Official leaderboard
Harbor-Index 4.90% #7 of 9 Official leaderboard
Chess Puzzles 19% #65 of 83 Epoch AI
Pencil Puzzle Bench 0% #52 of 53 Official leaderboard
CEO-Bench $0 #19 of 19 Official leaderboard
EdgeBench 31.0 #5 of 5 Official leaderboard

Professional work

BenchmarkScoreRankSource
Vals Finance Agent 60.4% #4 of 49 Official leaderboard
EnterpriseClawBench 58.8% #4 of 9 Official leaderboard
CommerceAgentBench 50.5% #6 of 13 Official leaderboard
CorpFin v2 61.4% #55 of 119 Official leaderboard
FORTE 70.5% #10 of 21 Official leaderboard
ProfBench 48.2% #35 of 72 Official leaderboard
GDPval-AA 1,554 #8 of 16 Model card
TaxEval v2 72.1% #65 of 129 Official leaderboard
Tax Agent Bench 25.2% #40 of 79 Official leaderboard
CaseLaw v2 59.4% #27 of 53 Official leaderboard
Harvey LAB (Legal Agent Benchmark) 3.75% #41 of 80 Official leaderboard
Public Benefits Bench 62.9% #26 of 50 Official leaderboard
$OneMillion-Bench 26% #7 of 13 Official leaderboard
PhysicianBench 18.7% #7 of 13 Official leaderboard
Clipping Bench $462 #11 of 19 Official leaderboard
LegalBench 80.3% #84 of 138 Official leaderboard
MedCode 40.5% #61 of 96 Official leaderboard
E-Commerce Bench 1.90 #13 of 19 Official leaderboard
Vending-Bench 2 $3,285 #47 of 66 Official leaderboard
Legal Research Bench 23.1% #63 of 88 Official leaderboard
Excel Modeling Benchmark 51.6% #55 of 76 Official leaderboard
Vals Finance Agent v2 44.1% #58 of 80 Official leaderboard
MedScribe 75.1% #74 of 98 Official leaderboard
PolicyBench 79.3% #39 of 50 Official leaderboard
prinzbench 23.0 #21 of 26 Official leaderboard
CAD Arena 0.19 #10 of 12 Official leaderboard
Ivo-micro1 ContractBench 48.4% #6 of 7 Official leaderboard
FrontierFinance 40.5% #14 of 16 Official leaderboard
Legora BAR 93.0 #16 of 17 Official leaderboard
DeepSeek Chinese white-collar tasks evaluation 53% #1 of 1 Model card

Vision, audio and robotics

BenchmarkScoreRankSource
ArtifactArena 765 #20 of 21 Official leaderboard
VoxelBench 1,241 #54 of 56 Official leaderboard
Code2Video Benchmark 1,387 #19 of 19 Official leaderboard

Safety and security

BenchmarkScoreRankSource
CyberGym 95.4% #4 of 33 Official leaderboard
PCT (Polarized Contrastive Pairs) 44% #22 of 46 Official leaderboard
Human Pathogen Capabilities Test v2 (HPCT v2) 52.7% #38 of 71 Official leaderboard
CyberBench (Vals) 68.6% #15 of 26 Official leaderboard
KORA 30.5% #26 of 42 Official leaderboard
DrugDiscoveryBench 31.7% #14 of 19 Official leaderboard
CheatBench 75.1% #8 of 10 Official leaderboard
Last Secure Code Benchmark 18% #5 of 6 Official leaderboard
Dictatorship Eval 11.7% #18 of 20 Official leaderboard

Other

BenchmarkScoreRankSource
ACCoRD 70.9 #45 of 150 Official leaderboard
Conceptual Reasoning Index 55.8 #48 of 149 Official leaderboard
LMCA 41.2% #49 of 139 Epoch AI
Political Science LLM Benchmark 66.3% #14 of 29 Official leaderboard
CAIS Text Capabilities Index 26.4 #39 of 66 Official leaderboard
SuperCLUE 64.4 #11 of 18 Official leaderboard
CAIS Risk Index 65.5 #58 of 65 Official leaderboard

DeepSeek's models over time

DeepSeek's models in color, against every other model on the Epoch Capabilities Index, by release date.

50 75 100 125 150 175 2023 2024 2025 2026 2027 DeepSeek V4.1 Flash 1017 DeepSeek-V4-Flash DeepSeek-V3.1 DeepSeek-V3 DeepSeek-V2 (MoE-236B, May 2024) DeepSeek Coder 1.3B DeepSeek Coder 33B 50 75 100 125 150 175 2023 2024 2025 2026 2027 DeepSeek V4.1 Flash 1017 DeepSeek-V4-Flash

Its launch post

The benchmarks the lab chose to headline, with the scores it gave.

DeepSeek-V4-Pro model card 24 Apr 2026

Its system card

Every result in the longer documents the lab published with it. 9 of them aren't in the launch post.

DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence (arXiv 2606.19348) 26 Apr 2026

Capabilities

Hallucination