Release dates
— DeepSeekMoE 16B was released publicly, exploring efficient MoE through fine-grained and shared-expert routing. ↗
DeepSeekMoE 16B
A 16.4B MoE using fine-grained expert segmentation and shared-expert isolation, released in Base and Chat variants.
MODEL NUMBER023
- First public date
- 2024-01-11
- Organization
- DeepSeek
- Family
- DeepSeek V / R
01
Specifications
Only published specifications are shown.
- Total parameters
- 16.4B
- Architecture
- MoE
- Context
- 4,096 tokens
- Input modalities
- Text
- Open weights
- Yes
- License
- DeepSeek Model License
- Status
- Legacy
02
Overview
Why it mattered
DeepSeekMoE was the key intermediate step validating DeepSeek’s efficient sparse-expert path before V2.
03
Architecture & capabilities
Model capabilities
Mixture of expertsMultilingual
Unclassified tags
Efficiency
04
Release history
DeepSeekMoE 16B was released publicly, exploring efficient MoE through fine-grained and shared-expert routing.
View source05
Related models
Select a card to view the related model.
06
Sources
Suggest a correction
Copy this template, add the field, proposed value, and primary source, and share it with the archive maintainer.