Skip to content
AI Atlas
ModelActiveOpen weights

Updated 12 min ago · first seen 11 Sept 2026

model_01M294X78JFS5FJEWGDRHPJBVS

Parameters
370.9B
T2 · 21 min ago
Released
21 May 2025
T2 · 16 min ago
License
mit
T2 · 16 min ago

Specification

Release date
21 May 2025

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium

Openness
open-weights

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 21 min agomedium

License
mit

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium

Architecture
DeepseekV3ForCausalLM

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium

Parameters
370.9B

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 21 min agomedium

Base model
deepseek-ai/DeepSeek-R1

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium

File size
386.3 GB

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium

Hugging Face repo
amd/DeepSeek-R1-MXFP4

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 21 min agomedium

Downloads
179,066

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 12 min agomedium

Likes
5

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 12 min agomedium

Gated
No

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium

Last modified
2026-04-13T16:42:07+00:00

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 21 min agomedium

Downloads all time
908,450

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 12 min agomedium

Model type
deepseek_v3

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium

Tags
safetensors, deepseek_v3, custom_code, 8-bit, quark

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium

Weights dtype
BF16, F32, U8

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium

Each value shows its source, tier and observation time. Conflicting claims are kept side by side and flagged — never averaged. How AI Atlas records facts →

Provenance

Attributed facts

17

Source tiers

T217

Freshest observation

12 min ago

Conflicts

None