Deepseek Logo

DeepSeek-MoE

Mixture of Experts model; efficient inference; reduced memory and computational requirements

Deepseek
Introduced in January 2024, DeepSeek-MoE represented a significant architectural shift toward efficiency using the Mixture of Experts (MoE) approach. The model featured specialized expert modules that activate selectively during inference, reducing computational overhead while maintaining or improving performance. Released in base and chat variants, DeepSeek-MoE demonstrated that sparse computation could deliver competitive performance with substantially lower training and inference costs. This innovation proved pivotal for DeepSeek's strategy of achieving high performance at reduced expense, influencing all subsequent model architectures including V2, V3, and beyond.

Reviews

No Reviews Yet

Be the first to share your experience with this AI tool

More models from Deepseek

DeepSeek-V3.1-Terminus Logo

DeepSeek-V3.1-Terminus

V3.1 refinement (September 2025); improved language consistency; enhanced agent performance; more stable outputs

Deepseek
DeepSeek-V3.2-Exp Logo

DeepSeek-V3.2-Exp

Experimental model (September 29, 2025); DeepSeek Sparse Attention (DSA) mechanism; 50-75% lower inference costs; long-context optimization

Deepseek
DeepSeek-V3.2 Logo

DeepSeek-V3.2

Production release (December 1, 2025); 671B parameters; DeepSeek Sparse Attention; GPT-5-level reasoning; 128K context

Deepseek
DeepSeek-V3.1 Logo

DeepSeek-V3.1

Hybrid model (August 2025); 671B parameters; dual-mode (thinking + non-thinking); 128K context; enhanced tool calling

Deepseek
DeepSeek-V3.2-Speciale Logo

DeepSeek-V3.2-Speciale

Extended reasoning variant (December 1, 2025); extreme thinking mode; 96% AIME score; gold IMO 2025; outperforms GPT-5-High

Deepseek
DeepSeek-Math Logo

DeepSeek-Math

Specialized mathematics model; three variants (Base, Instruct, RL); optimized for STEM problem-solving

Deepseek
DeepSeek-V2 Logo

DeepSeek-V2

Second generation MoE model; 236B total parameters (21B active); 128K context window; 50% lower training cost

Deepseek
DeepSeek-Coder-V2 Logo

DeepSeek-Coder-V2

Advanced coding model; 236B parameters (21B active); 128K context; 338 programming languages; GPT-4-Turbo-level coding

Deepseek
DeepSeek-V2.5 Logo

DeepSeek-V2.5

Refinement of V2; improved training data; enhanced transformer architecture; increased computational power

Deepseek
DeepSeek-Coder Logo

DeepSeek-Coder

First commercial-grade coding model; 1.3B-33B parameters; supports 80+ programming languages

Deepseek
DeepSeek-LLM Logo

DeepSeek-LLM

General-purpose foundation model; 67B parameters; outperforms LLaMA-2-70B on reasoning and math

Deepseek
DeepSeek-V3 Logo

DeepSeek-V3

Third generation foundation; 671B parameters (37B active); 128K context; major leap in capability and reasoning

Deepseek