LF

Lfm2-8B-A1B

LfmGGUF by unslothUpdated 2026-08-05
Download on Hugging Face
Parameters8.34B
Context125K
Layers24
Architecturelfm2moe
Smallest quant3.1 GB
Quick answer

Lfm2-8B-A1B needs about 7.2 GB of memory at Q4_K_M (a 5.0 GB file) with an 8K context window, so it fits on a 8 GB GPU. This repository has 19 quants, from 3.1 GB (Q2_K) to 16.7 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 6.4 GB.

LFM2-8B-A1B – a hybrid MoE model optimized for on‑device use

LFM2-8B-A1B is a 8.34 B parameter hybrid model with 1 B active parameters. It is designed for edge AI and on‑device deployment, offering quality comparable to 3‑4 B dense models while remaining lightweight. The model supports English, Arabic, Chinese, French, German, Japanese, Korean, and Spanish.

Good at

  • Agentic tasks
  • Data extraction
  • RAG and creative writing

Limits

  • Not suited for knowledge‑intensive or programming tasks
  • Requires 8 k context for memory estimates
  • License type is other

Best for

  • Fine‑tuning narrow use cases
  • Multi‑turn conversations
  • Tool‑use workflows

Who it’s for. Developers building lightweight agents or RAG systems for mobile or embedded devices. The model is ideal for teams that need efficient inference with limited GPU memory.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K3.1 GB 5.2 GB
Fits Get file →
Q2_K_L3.1 GB 5.2 GB
Fits Get file →
Q2_K_XL3.1 GB 5.2 GB
Fits Get file →
Q3_K_S3.6 GB 5.8 GB
Fits Get file →
Q3_K_XL3.7 GB 5.8 GB
Fits Get file →
Q3_K_M4.0 GB 6.1 GB
Fits Get file →
Q4_04.7 GB 6.8 GB
Fits Get file →
Q4_K_XL4.7 GB 6.9 GB
Fits Get file →
Q4_K_S4.8 GB 6.9 GB
Fits Get file →
Q4_K_M5.0 GB 7.2 GB
Fits Get file →
Q4_15.2 GB 7.4 GB
Tight Get file →
Q5_K_S5.8 GB 7.9 GB
Tight Get file →
Q5_K_M5.9 GB 8.0 GB
Too big Get file →
Q5_K_XL5.9 GB 8.0 GB
Too big Get file →
Q6_K6.8 GB 9.0 GB
Too big Get file →
Q6_K_XL6.9 GB 9.0 GB
Too big Get file →
Q8_08.9 GB 11.0 GB
Too big Get file →
Q8_K_XL9.0 GB 11.1 GB
Too big Get file →
BF1616.7 GB 18.8 GB
Too big Get file →

For 8 GB VRAM, use Q4_0 (file size 4.73 GB, estimated 6.8 GB memory). For 12 GB VRAM, Q5_K_M (5.92 GB, 8 GB memory) or Q6_K (6.85 GB, 9 GB memory) are suitable. Higher precision quants like BF16 require 18.8 GB memory and 16.69 GB file size.

What GPU do you need for Lfm2-8B-A1B?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: no quant fits at an 8K context.
  • 6 GB VRAM: the best quant that fits is Q3_K_XL (3.7 GB file, about 5.8 GB in use).
  • 8 GB VRAM: the best quant that fits is Q5_K_M (5.9 GB file, about 8.0 GB in use).
  • 12 GB VRAM: the best quant that fits is Q8_K_XL (9.0 GB file, about 11.1 GB in use).
  • 16 GB VRAM: the best quant that fits is Q8_K_XL (9.0 GB file, about 11.1 GB in use).
  • 24 GB VRAM: the best quant that fits is BF16 (16.7 GB file, about 18.8 GB in use).

Which Lfm2-8B-A1B quant should you choose?

Q4_K_M is the usual starting point for Lfm2-8B-A1B: it balances file size and quality well. Here it is a 5.0 GB file.

The Q8_0 file is 8.9 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

If memory is tight, the smallest file is Q2_K at 3.1 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturelfm2moe
Parameters8.34B (approximate, mixture of experts)
Active parameters1B
Layers24
Trained context128,000 tokens
KV cache (fp16)192 KB per token

Source

Quantized byunsloth
FamilyLfm
Licenseother
Downloads (HF)3,364
Last updated2026-08-05

Frequently asked questions about Lfm2-8B-A1B

How much VRAM does Lfm2-8B-A1B need?

About 7.2 GB at Q4_K_M with an 8K context window. That is the 5.0 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run Lfm2-8B-A1B on an 8 GB GPU?

Yes, with Q5_K_M: the 5.9 GB file needs about 8.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Lfm2-8B-A1B GGUF?

It depends on the quant: Q2_K is 3.1 GB, Q2_K_L is 3.1 GB, Q2_K_XL is 3.1 GB, Q3_K_S is 3.6 GB, Q3_K_XL is 3.7 GB, Q3_K_M is 4.0 GB and 13 more.

What context length does Lfm2-8B-A1B support?

The GGUF header lists a trained context length of 128,000 tokens (125K). The KV cache grows with the context you use: about 1.6 GB at 8K and 6.4 GB at 32K.

Where can I download Lfm2-8B-A1B GGUF?

The files are on Hugging Face in the unsloth/LFM2-8B-A1B-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Lfm2-8B-A1B use?

Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run LFM2-8B-A1B on 8 GB VRAM?

Yes, the Q4_0 quant uses about 6.8 GB memory at 8 k context, fitting within 8 GB VRAM.

What context length does the model support?

The model was trained with 128000 tokens and can operate with 8 k context for the provided memory estimates.

Is the model multilingual?

It supports English, Arabic, Chinese, French, German, Japanese, Korean, and Spanish.

Other GGUFs of LFM2-8B-A1B

Similar models

Browse all →

More from unsloth

All by unsloth →

More Lfm models

All Lfm →

Other small models

All small models →