LLM EXPLORER 61,249 MODELS INDEXED

DeepSeek V4.1 Flash EXL3 2.9bpw by Mia-AiLab

By Mia-AiLab · 763 downloads

DeepSeek V4.1 Flash EXL3 2.9bpw is an open-source language model by Mia-AiLab. Features: 105.2b LLM, VRAM: 209.7GB, License: mit, LLM Explorer Score: 0.36.

Base model:deepseek-ai/deepsee... Base model:quantized:deepseek-...   Conversational   Deepseek   Deepseek v41   Exl3   Exllamav3   Image-text-to-text   Moe   Multimodal   Quantized   Region:us   Safetensors   Sharded   Tensorflow   Vllm

DeepSeek V4.1 Flash EXL3 2.9bpw Parameters and Internals

LLM NameDeepSeek V4.1 Flash EXL3 2.9bpw
Repository 🤗https://huggingface.co/Mia-AiLab/DeepSeek-V4.1-Flash-EXL3-2.9bpw 
Base Model(s)  deepseek-ai/DeepSeek-V4.1-Flash   deepseek-ai/DeepSeek-V4.1-Flash
Model Size105.2b
Required VRAM209.7 GB
Updated2026-09-15
MaintainerMia-AiLab
Model Typedeepseek_v41
Model Files  6.5 GB: 1-of-39   5.3 GB: 2-of-39   5.2 GB: 3-of-39   5.2 GB: 4-of-39   5.2 GB: 5-of-39   5.2 GB: 6-of-39   5.2 GB: 7-of-39   5.2 GB: 8-of-39   5.2 GB: 9-of-39   5.2 GB: 10-of-39   5.2 GB: 11-of-39   5.2 GB: 12-of-39   5.2 GB: 13-of-39   5.2 GB: 14-of-39   5.3 GB: 15-of-39   5.2 GB: 16-of-39   5.2 GB: 17-of-39   5.2 GB: 18-of-39   7.0 GB: 19-of-39   7.0 GB: 20-of-39   3.5 GB: 21-of-39   5.2 GB: 22-of-39   5.2 GB: 23-of-39   5.2 GB: 24-of-39   5.2 GB: 25-of-39   5.2 GB: 26-of-39   5.2 GB: 27-of-39   5.2 GB: 28-of-39   5.2 GB: 29-of-39   5.2 GB: 30-of-39   5.2 GB: 31-of-39   5.2 GB: 32-of-39   5.2 GB: 33-of-39   5.2 GB: 34-of-39   5.2 GB: 35-of-39   5.2 GB: 36-of-39   5.2 GB: 37-of-39   8.1 GB: 38-of-39   5.8 GB: 39-of-39
Model ArchitectureDeepseekV41ForCausalLM
Licensemit
Model Max Length1048576
Transformers Version5.6.0
Tokenizer ClassPreTrainedTokenizerFast
Beginning of Sentence Token<|begin▁of▁sentence|>
End of Sentence Token<|end▁of▁sentence|>