LLM EXPLORER 60,958 MODELS INDEXED

DeepSeek V4.1 Flash REAP 2bit MLX by rapid-mlx

By rapid-mlx · 1114 downloads
HF Repo ↗ Copy URL

DeepSeek V4.1 Flash REAP 2bit MLX is an open-source language model by rapid-mlx. Features: 680.5b LLM, VRAM: 213.8GB, License: mit, Quantized.

  2-bit   2bit   Apple-silicon Base model:deepseek-ai/deepsee... Base model:quantized:deepseek-...   Deepseek-v41   Deepseek v41   Experimental   Mlx   Quantized   Reap   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/rapid-mlx/DeepSeek-V4.1-Flash-REAP-2bit-MLX 

DeepSeek V4.1 Flash REAP 2bit MLX Parameters and Internals

LLM NameDeepSeek V4.1 Flash REAP 2bit MLX
Repository 🤗https://huggingface.co/rapid-mlx/DeepSeek-V4.1-Flash-REAP-2bit-MLX 
Base Model(s)  deepseek-ai/DeepSeek-V4.1-Flash   deepseek-ai/DeepSeek-V4.1-Flash
Model Size680.5b
Required VRAM213.8 GB
Updated2026-09-13
Maintainerrapid-mlx
Model Typedeepseek_v41
Model Files  3.8 GB: 1-of-41   34.5 GB: 2-of-41   3.8 GB: 3-of-41   3.8 GB: 4-of-41   3.8 GB: 5-of-41   3.8 GB: 6-of-41   3.8 GB: 7-of-41   3.8 GB: 8-of-41   3.8 GB: 9-of-41   3.8 GB: 10-of-41   3.8 GB: 11-of-41   3.8 GB: 12-of-41   3.8 GB: 13-of-41   3.8 GB: 14-of-41   34.5 GB: 15-of-41   3.8 GB: 16-of-41   3.8 GB: 17-of-41   3.8 GB: 18-of-41   3.8 GB: 19-of-41   3.8 GB: 20-of-41   3.8 GB: 21-of-41   3.8 GB: 22-of-41   3.8 GB: 23-of-41   3.8 GB: 24-of-41   3.8 GB: 25-of-41   3.8 GB: 26-of-41   3.8 GB: 27-of-41   3.8 GB: 28-of-41   3.8 GB: 29-of-41   3.8 GB: 30-of-41   3.8 GB: 31-of-41   3.8 GB: 32-of-41   3.8 GB: 33-of-41   3.8 GB: 34-of-41   3.8 GB: 35-of-41   3.8 GB: 36-of-41   3.8 GB: 37-of-41   3.8 GB: 38-of-41   3.8 GB: 39-of-41   3.8 GB: 40-of-41   0.4 GB: 41-of-41
Quantization Type2bit
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｜>