LLM EXPLORER 60,384 MODELS INDEXED

DeepSeek V4 Flash Bf16 by mlx-community

By mlx-community · 400 downloads
HF Repo ↗ Copy URL

DeepSeek V4 Flash Bf16 is an open-source language model by mlx-community. Features: 284.3b LLM, VRAM: 162.3GB, Context: 1024K, LLM Explorer Score: 0.25.

  4-bit   Conversational   Deepseek v4   En   Mlx   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/mlx-community/DeepSeek-V4-Flash-bf16 

DeepSeek V4 Flash Bf16 Parameters and Internals

LLM NameDeepSeek V4 Flash Bf16
Repository 🤗https://huggingface.co/mlx-community/DeepSeek-V4-Flash-bf16 
Model Size284.3b
Required VRAM162.3 GB
Updated2026-08-10
Maintainermlx-community
Model Typedeepseek_v4
Model Files  5.0 GB: 1-of-33   4.9 GB: 2-of-33   4.8 GB: 3-of-33   5.2 GB: 4-of-33   4.9 GB: 5-of-33   4.8 GB: 6-of-33   5.2 GB: 7-of-33   4.9 GB: 8-of-33   4.8 GB: 9-of-33   5.2 GB: 10-of-33   4.9 GB: 11-of-33   4.8 GB: 12-of-33   5.2 GB: 13-of-33   4.9 GB: 14-of-33   4.8 GB: 15-of-33   5.2 GB: 16-of-33   4.9 GB: 17-of-33   4.8 GB: 18-of-33   5.2 GB: 19-of-33   4.9 GB: 20-of-33   4.8 GB: 21-of-33   5.2 GB: 22-of-33   4.9 GB: 23-of-33   4.8 GB: 24-of-33   5.2 GB: 25-of-33   4.9 GB: 26-of-33   4.8 GB: 27-of-33   5.2 GB: 28-of-33   4.9 GB: 29-of-33   4.8 GB: 30-of-33   5.2 GB: 31-of-33   4.9 GB: 32-of-33   3.4 GB: 33-of-33
Supported Languagesen
Model ArchitectureDeepseekV4ForCausalLM
Context Length1048576
Model Max Length1048576
Transformers Version4.57.1
Tokenizer ClassTokenizersBackend
Padding Token<｜end▁of▁sentence｜>
Vocabulary Size129280
Torch Data Typebfloat16

Best Alternatives to DeepSeek V4 Flash Bf16

Best Alternatives
Context / RAM
Downloads
Likes
DeepSeek V4 Flash1024K / 154.8 GB35321
DeepSeek V4 Flash BF161024K / 568.7 GB3608
DeepSeek V4 Flash Mxfp81024K / 154.8 GB8082
DeepSeek V4 Flash Mxfp41024K / 151.1 GB3661
DeepSeek V4 Flash 4bit1024K / 152.2 GB1263525
DeepSeek V4 Flash 2bit M DQ1024K / 106.4 GB7852
DeepSeek V4 Flash 3bit DQ1024K / 128.9 GB6962
...seek Ai DeepSeek V4 Flash 3bit1024K / 124.3 GB4771
...seek Ai DeepSeek V4 Flash 6bit1024K / 230.9 GB3671
DeepSeek V4 Flash 2bit DQ1024K / 96.3 GB1442547
Note: green Score (e.g. "73.2") means that the model is better than mlx-community/DeepSeek-V4-Flash-bf16.