llmboard.aiLeaderboard Center
Overall
Overall RankingOpen Models
Tools
Model DirectoryCompare Models
Capabilities
CodingReasoningMathKnowledgeInstruction Following
Price & Efficiency
Price & ValueCapability vs. PriceRuntime Performance
Modalities
Image GenerationVideo GenerationSpeech ModelsEmbeddings
Core Benchmarks
GPQAMMLU-ProAIME 2025SWE-Bench VerifiedMMLUHumanity's Last ExamLiveCodeBenchMATHHumanEvalMMMU-ProView all benchmarks
Methods
Scoring & Data
393 models668 benchmarks

Leaderboard Center

Overall RankingCodingCore BenchmarksPrice & ValueRuntime Performance

Modalities

All ModelsImage GenerationImage EditingVideo GenerationImage-to-VideoVideo EditingText-to-SpeechSpeech-to-TextEmbeddings

Data & Methods

Scoring MethodAll BenchmarksReasoningMath

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai

Model vendor

DeepSeek

Chinese AI company developing open and frontier language models, including the DeepSeek-V3 and DeepSeek-R1 model families.

Updated Aug 11, 2026. Data as of 2026-08-11

Website
Models10
Versions26
Open defaults3
CountryCN

On this page

  • Models
  • Scores
  • Highlights
  • About
  • FAQ

DeepSeek models

Default-version specifications and scored-version capability values are labeled separately.

10 rows
Columns

Show columns

DEDeepSeek-V4-FlashDeepSeekDeepSeek-V4-Flash-073178.320.0%1MOpen weightsJul 31, 2026DeepSeek-V4-Flash-0731
DEDeepSeek-V4-ProDeepSeekDeepSeek V4 Pro73.880.0%1MOpen weightsApr 24, 2026DeepSeek-V4-Pro-Max
DEDeepSeek-V3.2DeepSeekDeepSeek-V3.2-Speciale58.160.0%131.1KClosed weightsDec 1, 2025DeepSeek-V3.2-Speciale
DEDeepSeek-R1DeepSeekDeepSeek-R1-052844.280.0%131.1KClosed weightsMay 28, 2025DeepSeek-R1-0528
DEDeepSeek-V3.1DeepSeekDeepSeek-V3.139.980.0%131.1KOpen weightsJan 10, 2025DeepSeek-V3.1
DEDeepSeek-V3DeepSeekDeepSeek-V3 032426.080.0%163.8KClosed weightsMar 25, 2025DeepSeek-V3 0324
DEDeepSeek-R1-Distill-QwenDeepSeekDeepSeek R1 Distill Qwen 7B15.760.0%N/AClosed weightsJan 20, 2025DeepSeek R1 Distill Qwen 7B
DEDeepSeek-V2.5DeepSeekDeepSeek-V2.515.4100.0%8.2KClosed weightsMay 8, 2024DeepSeek-V2.5
DEDeepSeek-R1-Distill-LlamaDeepSeekDeepSeek R1 Distill Llama 8B12.460.0%N/AClosed weightsJan 20, 2025DeepSeek R1 Distill Llama 8B
DEDeepSeek-VL2DeepSeekDeepSeek VL2 Tiny0.00.0%N/AClosed weightsDec 13, 2024DeepSeek VL2 Tiny

DeepSeek capability scores

Each unique model uses its latest scored version.

DeepSeek LLMBoard scores

DeepSeek Highlights

Leading DeepSeek models by score, price and context window.

Top scoreDeepSeek-V4-Flash78.3 LLMBoardLowest official inputDeepSeek-V4-Flash$0.14 / 1MLargest contextDeepSeek-V4-Flash1M tokens
Open-weight defaults3of 10 models

What is DeepSeek?

Background and model catalog information for DeepSeek.

Chinese AI company developing open and frontier language models, including the DeepSeek-V3 and DeepSeek-R1 model families.

Model ownership is separate from API hosting. One model can be offered by multiple API providers with different prices.

Vendor
DeepSeek
Models
10
Versions
26
Country
CN
Founded
Not available

Catalog data as of 2026-08-11.

FAQ

Common questions about DeepSeek.

What is DeepSeek's highest-ranked model?

DeepSeek-V4-Flash currently has the strongest LLMBoard score from this vendor.

How many DeepSeek models are tracked?

The catalog contains 10 independently selectable DeepSeek models and 26 versions.

Does DeepSeek host every listed model API?

Not necessarily. Vendor identifies the model developer; API providers are tracked separately.

Does DeepSeek publish open-weight models?

3 current default versions are marked as open weight.