LLM EXPLORER 58,975 MODELS INDEXED

LFM2 8B A1B by LiquidAI

By LiquidAI · 30570 downloads
HF Repo ↗ Copy URL

LFM2 8B A1B is an open-source language model by LiquidAI. Features: 8b LLM, VRAM: 16.7GB, Context: 125K, License: other, LLM Explorer Score: 0.29.

  Arxiv:2511.23404   Ar   Conversational   De   Deploy:azure   Edge   En   Endpoints compatible   Es   Fr   Ja   Ko   Lfm2   Lfm2 moe   Liquid   Moe   Region:us   Safetensors   Sharded   Tensorflow   Zh
Model Card on HF 🤗: https://huggingface.co/LiquidAI/LFM2-8B-A1B 

LFM2 8B A1B Parameters and Internals

LLM NameLFM2 8B A1B
Repository 🤗https://huggingface.co/LiquidAI/LFM2-8B-A1B 
Model Size8b
Required VRAM16.7 GB
Updated2026-07-31
MaintainerLiquidAI
Model Typelfm2_moe
Model Files  5.0 GB: 1-of-4   5.0 GB: 2-of-4   5.0 GB: 3-of-4   1.7 GB: 4-of-4
Supported Languagesen ar zh fr de ja ko es
Model ArchitectureLfm2MoeForCausalLM
Licenseother
Context Length128000
Model Max Length128000
Transformers Version4.57.0.dev0
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<|pad|>
Vocabulary Size65536

Quantized Models of the LFM2 8B A1B

Model
Likes
Downloads
VRAM
LFM2 8B A1B 4bit91975 GB
LFM2 8B A1B 3bit MLX2974 GB
LFM2 8B A1B 8bit MLX7329 GB
LFM2 8B A1B 8bit3299 GB
LFM2 8B A1B 6bit MLX3207 GB
LFM2 8B A1B Fp1621433 GB

Best Alternatives to LFM2 8B A1B

Best Alternatives
Context / RAM
Downloads
Likes
LFM2.5 8B A1B125K / 16.9 GB84755670
...inal ToolBench Full SFT 1Epoch125K / 16.9 GB88672
LFM2.5 8B A1B Base125K / 16.9 GB420129
LFM2.5 8B A1B Opus Distil125K / 16.9 GB21275
LFM2.5 8B A1B125K / 16.9 GB124013
LFM2.5 8B A1B NVFP4125K / 6.9 GB4607
LFM2.5 8B A1B MLX Bf16125K / 16.8 GB3598
Supertron2.1 8B A1B125K / 17 GB966
... IQ1C Qwen3.6 35B A3B DISTILL2125K / 16.7 GB1113
LFM2.5 8B A1B ONNX125K /  GB7911
Note: green Score (e.g. "73.2") means that the model is better than LiquidAI/LFM2-8B-A1B.