LLM EXPLORER 58,744 MODELS INDEXED

LFM2.5 VL 3B Ffn8192 Int4 MLX Optimized by konic-labs

By konic-labs · 106 downloads
HF Repo ↗ Copy URL

LFM2.5 VL 3B Ffn8192 Int4 MLX Optimized is an open-source language model by konic-labs. Features: 3b LLM, VRAM: 1.6GB, License: apache-2.0, LLM Explorer Score: 0.3.

  Apple-silicon Base model:konic-labs/lfm2.5-v... Base model:quantized:konic-lab...   Compressed-tensors   Conversational   Custom code   En   Image-text-to-text   Lfm2   Lfm2 vl   Mlx   Mlx-vlm   Quantization   Region:us   Safetensors   Tool-calling   Vision   Vision-language-model
Model Card on HF 🤗: https://huggingface.co/konic-labs/LFM2.5-VL-3B-ffn8192-int4-MLX-optimized 

LFM2.5 VL 3B Ffn8192 Int4 MLX Optimized Parameters and Internals

LLM NameLFM2.5 VL 3B Ffn8192 Int4 MLX Optimized
Repository 🤗https://huggingface.co/konic-labs/LFM2.5-VL-3B-ffn8192-int4-MLX-optimized 
Base Model(s)  konic-labs/LFM2.5-VL-3B-ffn8192-int4   konic-labs/LFM2.5-VL-3B-ffn8192-int4
Model Size3b
Required VRAM1.6 GB
Updated2026-08-17
Maintainerkonic-labs
Model Typelfm2_vl
Model Files  1.6 GB
Supported Languagesen
Model ArchitectureLfm2VlForConditionalGeneration
Licenseapache-2.0
Transformers Version5.8.1
Tokenizer ClassTokenizersBackend
Padding Token<|pad|>

Best Alternatives to LFM2.5 VL 3B Ffn8192 Int4 MLX Optimized

Best Alternatives
Context / RAM
Downloads
Likes
LFM2.5 VL 3B0K / 6.2 GB073
LFM2.5 VL 3B MLX Bf160K / 6.2 GB21714
LFM2.5 VL 3B0K / 6.2 GB2592
...FM2.5 VL 3B MLX Mixed V6p6b4e40K / 2 GB682
LFM2.5 VL 3B ONNX0K /  GB1014
LFM2 VL 3B0K / 6 GB12342140
LFM2.5 VL 3B Ffn8192 Int40K / 2.6 GB100
LFM2.5 VL 3B Heretic0K / 6.2 GB01
LFM2.5 VL 3B MLX 4bit0K / 2.4 GB27215
LFM2.5 VL 3B MLX 8bit0K / 3.7 GB32413
Note: green Score (e.g. "73.2") means that the model is better than konic-labs/LFM2.5-VL-3B-ffn8192-int4-MLX-optimized.