llmboard.ai
Benchmarks
CompareRankings
llmboard.ai
Benchmarks
CompareRankings
HomeModelsDeepSeek-V3

DeepSeek model product

DeepSeek-V3

A powerful Mixture-of-Experts (MoE) language model with 671B total parameters (37B activated per token). Features Multi-head Latent Attention (MLA), auxiliary-loss-free load balancing, and multi-token prediction training. Pre-trained on 14.8T tokens with strong performance in reasoning, math, and code tasks.

Updated Aug 10, 2026. Default version: DeepSeek-V3 0324

Compare
LLMBoard score38.1DeepSeek-V3 0324
Coverage80%5 benchmark families
Context window163.8KTokens
Official input priceN/AOfficial price unavailable

On this page

  • Specification
  • Capability
  • Benchmarks
  • Arena
  • Pricing
  • Versions
  • About
  • Compare
  • Similar models
  • FAQ

Model specification

Structured fields from the published default version.

Version
DeepSeek-V3 0324
Released
Mar 25, 2025
Knowledge cutoff
Unknown
Parameters
671B
Context window
163.8K
Max output
163.8K
Inputs
text
Outputs
text
Open weights
No
License
MIT + Model License (Commercial use allowed)

Capability profile

This profile uses the latest version under this unique model that has a calculated LLMBoard score. Arena and price are excluded.

DeepSeek-V3 0324 category scores

Benchmark results

Published benchmark records for the scored version DeepSeek-V3 0324.

5 rows
Columns

Show columns

MATH-5000.9223232.3%CAug 7, 2026
AIME 20240.6445317.3%CAug 7, 2026
MMLU-Pro0.84412966.4%CAug 7, 2026
LiveCodeBench0.5477336.1%CAug 7, 2026
GPQA0.713523342.2%CAug 7, 2026

Arena results

Preference and agent-evaluation signals from published Arena datasets.

58 rows
Columns

Show columns

text style controlcreative writing1081389.96,031N/AAug 6, 2026
text style controlgerman1111398.41,154N/AAug 6, 2026
text style controlpolish1171395.32,689N/AAug 6, 2026
textjapanese1181326.8961N/AAug 6, 2026
text style controlindustry entertainment and sports and media1181382.68,044N/AAug 6, 2026
text style controljapanese1191337.4961N/AAug 6, 2026
text style controlfrench1221420.3539N/AAug 6, 2026
text style controlmulti turn1231410.37,850N/AAug 6, 2026
textgerman1241380.21,154N/AAug 6, 2026
textcreative writing1271364.56,031N/AAug 6, 2026
text style controlrussian1271399.03,094N/AAug 6, 2026
text style controlindustry writing and literature and language1281386.410,322N/AAug 6, 2026
text style controlindustry legal and government1311410.62,945N/AAug 6, 2026
textpolish1321367.92,689N/AAug 6, 2026
textfrench1331388.7539N/AAug 6, 2026
text style controlindustry medicine and healthcare1341427.12,662N/AAug 6, 2026
textindustry entertainment and sports and media1351356.98,044N/AAug 6, 2026
text style controlkorean1351333.8922N/AAug 6, 2026
textkorean1371316.6922N/AAug 6, 2026
text style controlindustry life and physical and social science1401414.47,671N/AAug 6, 2026
text style controlnon english1401380.122,199N/AAug 6, 2026
textmulti turn1411387.97,850N/AAug 6, 2026
text style controloverall1411395.645,445N/AAug 6, 2026
text style controlexclude ties1411376.631,818N/AAug 6, 2026
textindustry writing and literature and language1441364.910,322N/AAug 6, 2026
textrussian1441371.93,094N/AAug 6, 2026
text style controlenglish1451407.623,241N/AAug 6, 2026
text style controlindustry business and management and financial operations1471390.36,919N/AAug 6, 2026
text style controlchinese1481413.62,466N/AAug 6, 2026
textenglish1491386.623,241N/AAug 6, 2026
textindustry legal and government1501380.52,945N/AAug 6, 2026
textnon english1501360.522,199N/AAug 6, 2026
textoverall1511375.145,445N/AAug 6, 2026
text style controlindustry software and it services1511423.914,611N/AAug 6, 2026
text style controllonger query1511393.68,447N/AAug 6, 2026
textexclude ties1521346.031,818N/AAug 6, 2026
text style controlhard prompts1521408.718,671N/AAug 6, 2026
textindustry life and physical and social science1531381.77,671N/AAug 6, 2026
text style controlhard prompts english1531418.010,217N/AAug 6, 2026
text style controlinstruction following1541378.612,386N/AAug 6, 2026
textindustry medicine and healthcare1551382.42,662N/AAug 6, 2026
text style controlexpert1571396.92,295N/AAug 6, 2026
text style controlspanish1581364.3893N/AAug 6, 2026
textchinese1591388.12,466N/AAug 6, 2026
textmath1611374.03,183N/AAug 6, 2026
text style controlcoding1611428.88,358N/AAug 6, 2026
texthard prompts1621365.818,671N/AAug 6, 2026
textspanish1621345.9893N/AAug 6, 2026
textinstruction following1631345.612,386N/AAug 6, 2026
texthard prompts english1641376.110,217N/AAug 6, 2026
textindustry business and management and financial operations1641350.26,919N/AAug 6, 2026
textexpert1671351.52,295N/AAug 6, 2026
textindustry software and it services1671379.714,611N/AAug 6, 2026
textlonger query1681352.98,447N/AAug 6, 2026
textindustry mathematical1701365.32,923N/AAug 6, 2026
text style controlmath1711369.53,183N/AAug 6, 2026
textcoding1741368.98,358N/AAug 6, 2026
text style controlindustry mathematical1741367.32,923N/AAug 6, 2026

Pricing

Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.

Official API
N/A
Official provider
N/A
Lowest third-party
From $0.20 input, $0.77 output per 1M via NanoGPT
Tracked offerings
7
6 rows
Columns

Show columns

NanoGPTdeepseek-v3-0324global$0.20$0.77128KAug 7, 2026
submodeldeepseek-ai/DeepSeek-V3-0324global$0.20$0.8075KAug 7, 2026
Meganovadeepseek-ai/DeepSeek-V3-0324global$0.25$0.88163.8KAug 7, 2026
NovitaAIdeepseek/deepseek-v3-0324global$0.27$1.12163.8KAug 7, 2026
OpenRouterdeepseek/deepseek-chat-v3-0324global$0.27$1.12163.8KAug 7, 2026
Jiekou.AIdeepseek/deepseek-v3-0324global$0.28$1.14163.8KAug 7, 2026

Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.

DeepSeek-V3 versions

All published versions linked to this unique model. The score columns identify the version used by the current overall ranking.

2 rows
Columns

Show columns

DeepSeek-V3 0324Mar 25, 202538.1671B163.8K163.8KNoMIT + Model License (Commercial use allowed)
DeepSeek-V3Dec 25, 2024N/A671B131.1K8.2KYesMIT + Model License (Commercial use allowed)

What is DeepSeek-V3?

A concise description based on the published model registry.

A powerful Mixture-of-Experts (MoE) language model with 671B total parameters (37B activated per token). Features Multi-head Latent Attention (MLA), auxiliary-loss-free load balancing, and multi-token prediction training. Pre-trained on 14.8T tokens with strong performance in reasoning, math, and code tasks.

Use the benchmark, Arena and pricing sections above as separate evidence. A missing field means the current data snapshot does not support that claim.

Data snapshot: 2026-08-07. Editorial model content is not available in the backend.

DeepSeek-V3 vs nearby models

Open a comparison with the three ranked models immediately above and below this model.

DeepSeek-V3vsGemma 3 12BDeepSeek-V3vsJamba 1.5 LargeDeepSeek-V3vsDiffusionGemma 26B A4BDeepSeek-V3vsQwen2 72BDeepSeek-V3vsPhi 4DeepSeek-V3vsGemini 1.5 Flash

Models similar to DeepSeek-V3

Recommendations prioritize the same model type and family, then the closest published LLMBoard score.

#126-2.1
DE

DeepSeek-V3.1

DeepSeek

36.0 LLMBoard

DetailsCompare
#101+6.0
DE

DeepSeek-R1

DeepSeek

44.1 LLMBoard

DetailsCompare
#32+31.4
DE

DeepSeek-V2.5

DeepSeek

69.5 LLMBoard

DetailsCompare
#28+33.6
DE

DeepSeek-V4-Flash

DeepSeek

71.7 LLMBoard

DetailsCompare
#18+37.4
DE

DeepSeek-V3.2

DeepSeek

75.5 LLMBoard

DetailsCompare
#12+42.0
DE

DeepSeek-V4-Pro

DeepSeek

80.1 LLMBoard

DetailsCompare

FAQ

Common questions about DeepSeek-V3.

When was DeepSeek-V3 released?

DeepSeek-V3's default version was released on Mar 25, 2025.

How much does DeepSeek-V3 cost?

No official standard PAYG price is currently available for DeepSeek-V3. The lowest tracked third-party offer starts at $0.20 input and $0.77 output via NanoGPT.

Who created DeepSeek-V3?

DeepSeek-V3 is published under DeepSeek in the model registry.

What is the context window for DeepSeek-V3?

The default version has a 163.8K token context window.

Is DeepSeek-V3 open weight?

No. The default version is not marked as having publicly available weights.

How many API providers offer DeepSeek-V3?

7 published provider offerings are linked to the default version.

What models should I compare DeepSeek-V3 with?

Nearby ranked alternatives include Gemma 3 12B, Jamba 1.5 Large, DiffusionGemma 26B A4B.

Rankings

OverallCodingText ArenaPricing

Modalities

Image GenerationVideo GenerationSpeech-to-TextEmbeddings

Benchmarks

All BenchmarksReasoningMathCoding

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai