LFM2 24B A2B needs about 17.6 GB of memory at Q4_K_M (a 14.4 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 7 quants, from 13.5 GB (Q4_0) to 47.7 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 10.7 GB.
24B MoE LLM optimized for on‑device inference
LFM2‑24B‑A2B is a 24 billion‑parameter hybrid model with only 2 billion active parameters per token. It is designed for efficient edge deployment and supports function calling and retrieval‑augmented generation.
Good at
- Fast local inference
- Function calling
- Retrieval‑augmented generation
Limits
- Requires 32 GB RAM for full model
- No vision support
- License is other
Best for
- Agentic tool use
- Offline document summarization
- Privacy‑preserving support agents
Who it’s for. Developers building local LLM applications that need a large model with low memory footprint. Suitable for on‑premise customer support, legal or medical Q&A, and single‑machine RAG pipelines.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_0 | 13.5 GB | 16.7 GB | Too big | Get file → | |
| Q4_K_M | 14.4 GB | 17.6 GB | Too big | Get file → | |
| Q5_K_M | 16.9 GB | 20.1 GB | Too big | Get file → | |
| Q6_K | 19.6 GB | 22.8 GB | Too big | Get file → | |
| Q8_0 | 25.4 GB | 28.5 GB | Too big | Get file → | |
| BF16 | 47.7 GB | 50.9 GB | Too big | Get file → | |
| F16 | 47.7 GB | 50.9 GB | Too big | Get file → |
For 8 GB VRAM use Q4_0 (13.47 GB file, ~16.7 GB RAM at 8k context). For 16 GB VRAM Q4_K_M (14.42 GB, ~17.6 GB RAM). For 32 GB VRAM Q8_0 (25.35 GB, ~28.5 GB RAM).
What GPU do you need for LFM2 24B A2B?
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: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is Q6_K (19.6 GB file, about 22.8 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (25.4 GB file, about 28.5 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (25.4 GB file, about 28.5 GB in use).
Which LFM2 24B A2B quant should you choose?
Q4_K_M is the usual starting point for LFM2 24B A2B: it balances file size and quality well. Here it is a 14.4 GB file.
The Q8_0 file is 25.4 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 Q4_0 at 13.5 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about LFM2 24B A2B
How much VRAM does LFM2 24B A2B need?
About 17.6 GB at Q4_K_M with an 8K context window. That is the 14.4 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 24B A2B on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 16.7 GB, so you would need a larger GPU or CPU offloading.
What is the file size of LFM2 24B A2B GGUF?
It depends on the quant: Q4_0 is 13.5 GB, Q4_K_M is 14.4 GB, Q5_K_M is 16.9 GB, Q6_K is 19.6 GB, Q8_0 is 25.4 GB, BF16 is 47.7 GB and 1 more.
What context length does LFM2 24B A2B support?
The GGUF header lists a trained context length of 128,000 tokens (125K). The KV cache grows with the context you use: about 2.7 GB at 8K and 10.7 GB at 32K.
Where can I download LFM2 24B A2B GGUF?
The files are on Hugging Face in the LiquidAI/LFM2-24B-A2B-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 24B A2B 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 LFM2‑24B‑A2B on a laptop with 16 GB RAM?
Yes, the Q4_K_M quantized model fits within 16 GB RAM, but performance may be limited by CPU speed.
Does the model support image input?
No, LFM2‑24B‑A2B does not have vision capabilities.
What is the maximum context length?
The model was trained on 128,000 tokens and supports up to that context length.