DeepSeek V4 Pro 0813

  • Open weights
  • Input: text
  • Released 2026-08
RankedEnough results for an index.
IndexAll categories, out of 100.64.0±4.6Median 50.7
CoverageIndex weight that has results.85%
SpeedOutput tokens per second.93/sMedian 76/s
First tokenSeconds to the first output token.0.3 sMedian 1.1 s
Input / 1MUS dollars per 1M input tokens.$1.32batch $0.66Median $0.60
Output / 1MUS dollars per 1M output tokens.$3.96batch $1.98 Batch price per 1M output tokens.Median $2.50
ContextMaximum tokens in one request.1.05MMedian 262K
EloLMArena rating and rank.—Median 1425

Details

From OpenRouter
API IDThe model ID on OpenRouter.
deepseek/deepseek-v4-pro-0813
Max outputMaximum output tokens per request.
944K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.044
Weights
Hugging Face (opens in a new tab) ↗
RetiresDate OpenRouter removes the model.
—

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.4 counted58.579 / 221
CodingCode writing and repair.9 counted65.340 / 214
ReasoningLogic problems and puzzles.5 counted67.530 / 225
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.4 counted62.057 / 266
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.1 counted62.7—
MathMath problems.5 counted64.334 / 171

Results

28 counted
BenchmarkResultThe score from the source.PlacePlace among results on this benchmark.
SWE-benchCoding96.4% ±0.82 / 87
LiveCodeBenchCoding87.5% ±1.013 / 142
GPQA Diamond · Vals AIKnowledge92.4% ±2.016 / 135
OTIS Mock AIME 2024-2025Math98.6% ±1.030 / 203
GPQA diamond · Epoch AIKnowledge91.7% ±1.534 / 220
LiveBench MathematicsMath95.1%14 / 66
Vibe Code Bench v1.1Coding82.3% ±3.130 / 130
LiveBench Data AnalysisReasoning79.2%19 / 80
Mystery Game PuzzlesReasoning43.0% ±5.020 / 83
MMLU ProKnowledge87.0% ±0.335 / 135
ARC-AGI-1 (semi-private)Reasoning90.0%39 / 132
ARC-AGI-2 (semi-private)Reasoning61.3%42 / 137
SimpleQA VerifiedKnowledge52.9% ±1.629 / 94
Chess PuzzlesReasoning46.0% ±5.029 / 93
LiveBench LanguageKnowledge82.1%24 / 66
Code MigrationCoding41.5% ±4.328 / 77
LiveBench ReasoningReasoning85.8%35 / 80
LiveBench Agentic CodingAgentic54.9%29 / 66
SkillsBenchCoding53.8% ±4.516 / 36
FrontierMath-Tiers-1-3-v2-PrivateMath64.6% ±2.842 / 93
LiveBench Instruction FollowingInstruction67.7%34 / 66
IOICoding51.6% ±2.425 / 46
FrontierMath-Tier-4-v2-PrivateMath26.8% ±7.042 / 75
LiveBench CodingCoding77.2%37 / 66
ProofBench v1.1Math50.0% ±5.033 / 55
Vibe Code Bench 1-100Coding17.5% ±3.416 / 26
Terminal-Bench 4.0Agentic14.1% ±2.045 / 68
Terminal-Bench 2.1 · Vals AIAgentic54.7% ±1.551 / 75
MirrorCodeCoding33.3% ±10.029 / 40
Terminal-Bench 4.1.0Agentic6.7% ±1.346 / 52
CorpFin v2Finance65.4% ±0.927 / 131
CyberBench v1.1Other72.9% ±5.515 / 49
ARC-AGI-2 (public eval)Reasoning59.4%45 / 127
TaxEval v2Finance73.1% ±0.952 / 143
ARC-AGI-1 (public eval)Reasoning93.3%43 / 116
Legal Research BenchLegal40.9% ±3.438 / 98
MedCodeHealthcare42.5% ±2.251 / 108
Tax Agent BenchFinance26.8% ±3.143 / 89
LegalBenchLegal82.4% ±0.472 / 149
ProgramBenchCoding0.0%35 / 66
MedScribeHealthcare80.2% ±2.061 / 111
Finance Agent (v2)Finance50.4% ±0.244 / 79
HLAB-Vals-1.1Legal10.8% ±2.724 / 42
Terminal-Bench-ScienceScience4.3% ±2.429 / 48
SRE BenchOther1.9% ±0.825 / 37
EMBFinance52.8% ±3.153 / 75

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseEpoch AICC BYVals AINo licence statedLiveBenchNo licence statedARC Prize FoundationTerms restrict reuseTerminal-BenchNo licence statedmodels.devMITCheaper InferenceNo licence stated

History

13 releases
Index64.1 → 64.0
18 Nov 202621 Nov 2026
Output / 1M$4.44 → $3.96
18 Nov 202621 Nov 2026

Questions

How much does DeepSeek V4 Pro 0813 cost?
$1.32 per 1M input tokens, $3.96 per 1M output tokens.
How good is DeepSeek V4 Pro 0813?
Its index is 64.0 out of 100. That is place 50 of 282 ranked models.
What is DeepSeek V4 Pro 0813 best at?
Reasoning: 67.5 out of 100, place 30 of 225.
How fast is DeepSeek V4 Pro 0813?
93 output tokens per second, 0.3 s to the first token.
What is the context window of DeepSeek V4 Pro 0813?
1.05M tokens.
Is DeepSeek V4 Pro 0813 open source?
Yes. Its weights are published.
When was DeepSeek V4 Pro 0813 released?
August 2026.

Badge

DeepSeek V4 Pro 0813 on Inference360: place 50 of 282, index 64.0

Compare with

Muse Spark 1.263.9Muse Ember64.0Muse Spark Open64.1Muse Spark 1.163.6DeepSeek V4 Flash 073160.6

Lower price

dots3-note Preview62.3 · FreeGLM-5.3-Flash62.4 · $0.078/1MHy4 flash64.5 · $0.14/1M

More from DeepSeek

All models
DeepSeek V4 Pro60.3DeepSeek V4 Flash Vision Exp59.6DeepSeek V4 Flash LatestUnrankedDeepSeek V4 Flash 073160.6DeepSeek Flash LatestUnranked