LLM EXPLORER 58,975 MODELS INDEXED

LFM2.5 8B A1B MLX 6bit by LiquidAI

By LiquidAI · 439 downloads
HF Repo ↗ Copy URL

LFM2.5 8B A1B MLX 6bit is an open-source language model by LiquidAI. Features: 8b LLM, VRAM: 6.9GB, Context: 125K, License: other, Quantized, LLM Explorer Score: 0.26.

  Arxiv:2511.23404   6-bit   6bit   Ar Base model:liquidai/lfm2.5-8b-... Base model:quantized:liquidai/...   Conversational   De   Edge   En   Es   Fr   Ja   Ko   Lfm2.5   Lfm2 moe   Liquid   Mlx   Moe   Pt   Quantized   Region:us   Safetensors   Sharded   Tensorflow   Zh
Model Card on HF 🤗: https://huggingface.co/LiquidAI/LFM2.5-8B-A1B-MLX-6bit 

LFM2.5 8B A1B MLX 6bit Parameters and Internals

LLM NameLFM2.5 8B A1B MLX 6bit
Repository 🤗https://huggingface.co/LiquidAI/LFM2.5-8B-A1B-MLX-6bit 
Base Model(s)  LiquidAI/LFM2.5-8B-A1B   LiquidAI/LFM2.5-8B-A1B
Model Size8b
Required VRAM6.9 GB
Updated2026-08-08
MaintainerLiquidAI
Model Typelfm2_moe
Model Files  5.3 GB: 1-of-2   1.6 GB: 2-of-2
Supported Languagesen ar zh fr de ja ko es pt
Quantization Type6bit
Model ArchitectureLfm2MoeForCausalLM
Licenseother
Context Length128000
Model Max Length128000
Transformers Version5.2.0
Tokenizer ClassTokenizersBackend
Padding Token<|pad|>
Vocabulary Size128000

Best Alternatives to LFM2.5 8B A1B MLX 6bit

Best Alternatives
Context / RAM
Downloads
Likes
LFM2.5 8B A1B MLX 8bit125K / 9 GB151429
LFM2.5 8B A1B MLX 4bit125K / 4.8 GB122015
LFM2.5 8B A1B MLX 4bit125K / 4.8 GB5201
LFM2.5 8B A1B MLX 5bit125K / 5.9 GB3068
LFM2.5 8B A1B MLX 8bit125K / 9 GB2775
LFM2 8B A1B 4bit125K / 5.2 GB1979
LFM2 8B A1B 3bit MLX125K / 4.2 GB972
LFM2 8B A1B 8bit MLX125K / 9.4 GB327
LFM2 8B A1B 8bit125K / 9.4 GB293
LFM2 8B A1B 6bit MLX125K / 7.3 GB203
Note: green Score (e.g. "73.2") means that the model is better than LiquidAI/LFM2.5-8B-A1B-MLX-6bit.