LLM EXPLORER 61,249 MODELS INDEXED

DeepSeek V4.1 MLX Q4i by inferencerlabs

By inferencerlabs · 2054 downloads
HF Repo ↗ Copy URL

DeepSeek V4.1 MLX Q4i is an open-source language model by inferencerlabs. Features: LLM, VRAM: 144.9GB, LLM Explorer Score: 0.34.

Base model:deepseek-ai/deepsee... Base model:quantized:deepseek-...   Conversational   Deepseek v41   En   Image-text-to-text   Mlx   Quantized   Region:us   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/inferencerlabs/DeepSeek-V4.1-MLX-Q4i 

DeepSeek V4.1 MLX Q4i Parameters and Internals

LLM NameDeepSeek V4.1 MLX Q4i
Repository 🤗https://huggingface.co/inferencerlabs/DeepSeek-V4.1-MLX-Q4i 
Base Model(s)  deepseek-ai/DeepSeek-V4.1-Flash   deepseek-ai/DeepSeek-V4.1-Flash
Required VRAM144.9 GB
Updated2026-09-15
Maintainerinferencerlabs
Model Typedeepseek_v41
Model Files  5.3 GB: 1-of-123   4.6 GB: 2-of-123   4.3 GB: 3-of-123   6.1 GB: 4-of-123   0.2 GB: 5-of-123   6.1 GB: 6-of-123   0.2 GB: 7-of-123   6.1 GB: 8-of-123   0.2 GB: 9-of-123   6.1 GB: 10-of-123   0.2 GB: 11-of-123   6.1 GB: 12-of-123   0.2 GB: 13-of-123   6.1 GB: 14-of-123   0.2 GB: 15-of-123   6.1 GB: 16-of-123   0.2 GB: 17-of-123   6.1 GB: 18-of-123   0.2 GB: 19-of-123   6.1 GB: 20-of-123   0.2 GB: 21-of-123   6.1 GB: 22-of-123   0.2 GB: 23-of-123   6.1 GB: 24-of-123   0.2 GB: 25-of-123   6.1 GB: 26-of-123   0.2 GB: 27-of-123   6.1 GB: 28-of-123   0.2 GB: 29-of-123   6.1 GB: 30-of-123   0.2 GB: 31-of-123   6.1 GB: 32-of-123   0.2 GB: 33-of-123   6.1 GB: 34-of-123   4.5 GB: 35-of-123   4.4 GB: 36-of-123   4.4 GB: 37-of-123   4.0 GB: 38-of-123   4.4 GB: 39-of-123   4.4 GB: 40-of-123   4.0 GB: 41-of-123
Supported Languagesen
Model ArchitectureDeepseekV41ForCausalLM
Transformers Version5.6.0

Best Alternatives to DeepSeek V4.1 MLX Q4i

Best Alternatives
Context / RAM
Downloads
Likes
DeepSeek V4.1 Flash REAP 256E0K / 206.2 GB1046
DeepSeek V4.1 Flash REAP 272E0K / 218.2 GB721
...epSeek V4.1 Flash MixedQ2 GGUF0K / 10.8 GB68024
Note: green Score (e.g. "73.2") means that the model is better than inferencerlabs/DeepSeek-V4.1-MLX-Q4i.