Name: Deepseek MoE 16B Base
Author: deepseek-ai

Deepseek MoE 16B Base is an open-source language model by deepseek-ai. Features: 16.4b LLM, VRAM: 32.8GB, Context: 4K, License: other, MoE, HF Score: 51.1, LLM Explorer Score: 0.22, Arc: 53.2, HellaSwag: 79.8, MMLU: 46.3, TruthfulQA: 36.1, WinoGrande: 73.7, GSM8K: 17.3.

Arxiv:2401.06066 Custom code Deepseek Moe Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/deepseek-ai/deepseek-moe-16b-base

Deepseek MoE 16B Base Benchmarks

IFEval: 24.5 vs 88 (so35)^-72.2%

ARC: 53.24 vs 96.7 (so35)^-44.9%

HellaSwag: 79.77 vs 95.3 (gpt4)^-16.3%

MMLU: 46.31 vs 88.3 (so35)^-47.6%

TruthfulQA: 36.08 vs 59 (gpt4)^-38.8%

WinoGrande: 73.72 vs 87.5 (gpt4)^-15.7%

GSM8K: 17.29 vs 96.4 (so35)^-82.1%

MATH Lvl 5: 2.42

LLME Score: 0.21852

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

Deepseek MoE 16B Base (deepseek-ai/deepseek-moe-16b-base)

🌟 Advertise your project 🚀

Deepseek MoE 16B Base Parameters and Internals

LLM Name	Deepseek MoE 16B Base
Repository 🤗	https://huggingface.co/deepseek-ai/deepseek-moe-16b-base
Model Size	16.4b
Required VRAM	32.8 GB
Updated	2026-04-03
Maintainer	deepseek-ai
Model Type	deepseek
Model Files	5.0 GB: 1-of-7 5.0 GB: 2-of-7 5.0 GB: 3-of-7 5.0 GB: 4-of-7 5.0 GB: 5-of-7 5.0 GB: 6-of-7 2.8 GB: 7-of-7
Model Architecture	DeepseekForCausalLM
License	other
Context Length	4096
Model Max Length	4096
Transformers Version	4.36.0
Tokenizer Class	LlamaTokenizerFast
Beginning of Sentence Token	<｜begin▁of▁sentence｜>
End of Sentence Token	<｜end▁of▁sentence｜>
Vocabulary Size	102400
Torch Data Type	bfloat16

Best Alternatives to Deepseek MoE 16B Base

Best Alternatives	Context / RAM	Downloads	Likes
Deepseek MoE 16B Chat	4K / 32.8 GB	4434	150

Rank the Deepseek MoE 16B Base Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

What open-source LLMs or SLMs are you in search of? 52473 in total.

Email us: info@extractum.io. Our Privacy Policy | Terms and Conditions | Suggest an improvement.

Our Social Media →

Original data from HuggingFace, OpenCompass and various public git repos.

Check out Ag3ntum — our secure, self-hosted AI agent for server management.

Release v20260328a

Support LLM Explorer

Deepseek MoE 16B Base by deepseek-ai

» All LLMs » deepseek-ai » Deepseek MoE 16B Base URL Share it on

Deepseek MoE 16B Base Benchmarks

Deepseek MoE 16B Base Parameters and Internals

Best Alternatives to Deepseek MoE 16B Base

Rank the Deepseek MoE 16B Base Capabilities

What open-source LLMs or SLMs are you in search of? 52473 in total.