text model · LFM · macOS

Can I run LFM2 24B-A2B on Apple M3 Ultra (256GB)?

Compatibility verdict VRAM threshold engine

Yes, it runs GPU accelerated

Yes. LFM2 24B-A2B runs on Apple M3 Ultra (256GB) at Q4_K_M (~15.4 GB of ~192 GB usable).

Needs ~15.4 GB Device usable ~192 GB

Runs at Q4_K_M using ~15.4 GB of ~192 GB usable. You have room for FP16 for higher quality.

That figure is at a 4k context and moves about ±15% as context length changes. Apple M3 Ultra (256GB) leaves ~176.6 GB of headroom, room to step up to FP16 for higher quality.

Q4_K_M needed: ~15.4 GB
Usable on device: ~192 GB
Device memory: 256 GB
Best quant: Q4_K_M

Share on X Share on Reddit

Which quant fits

Quant ladder vs ~192 GB usable

Q2_K

~12.1 GB

Q3_K_M

~13.7 GB

Q4_K_M

~15.4 GB

Q5_K_M

~19.1 GB

Q6_K

~21.7 GB

Q8_0

~25.6 GB

FP16

~46.4 GB

The line marks Apple M3 Ultra (256GB)'s ~192 GB budget; rungs past it are too large.

Run it

Install commands macOS

Pick your tool. All 3 load the same Q4_K_M weights.

Ollama

$ ollama run lfm2:24b

llama.cpp

$ llama-cli -hf LiquidAI/LFM2-24B-A2B-GGUF:Q4_K_M

LM Studio

$ lms get LiquidAI/LFM2-24B-A2B-GGUF

Model LFM

Parameters: 24B (MoE, 2.3B active)
Q4_K_M size: 13.43 GB
Q8_0 size: 23.61 GB
Context: 32k
Ollama tag: lfm2:24b

Full LFM2 24B-A2B requirements →

Device macOS

Memory: 256 GB unified
Usable for weights: ~192 GB
Power draw: ~270 W
Best runtime: MLX direct / Ollama (MLX backend)

Best models for Apple M3 Ultra (256GB) →

You could also run

Run LFM2 24B-A2B on other hardware

Apple M4 (24GB)Apple M4 Pro (24GB)Nvidia GeForce RTX 4090 (24GB)Nvidia GeForce RTX 3090 (24GB)AMD Radeon RX 7900 XTX (24GB)Nvidia GeForce RTX 5090 (32GB)

FAQ

Can Apple M3 Ultra (256GB) run LFM2 24B-A2B?

Yes. LFM2 24B-A2B runs on Apple M3 Ultra (256GB) at Q4_K_M (~15.4 GB of ~192 GB usable).

How much memory does LFM2 24B-A2B need?

Apple M3 Ultra (256GB) has room to spare. At Q4_K_M the weights are ~13.43 GB; with KV cache and runtime overhead, budget ~15.4 GB at a 4k context. It is a Mixture-of-Experts model (24B total / 2.3B active), so all experts must stay in memory; memory tracks total params, not active params.

What is the best tool to run LFM2 24B-A2B on macOS?

LM Studio for a simple setup; mlx-lm for the most speed. vLLM is NOT a Mac tool, it is a CUDA/Linux serving engine. Unified memory is not a fixed VRAM slice; ~70% is usable for weights.

Embed this

A live badge for your model card or README, updated as the data is.

Markdown

[![LFM2 24B-A2B on Apple M3 Ultra (256GB)](https://localmodel.run/badge/lfm2-24b-a2b/apple-m3-ultra-256gb.svg)](https://localmodel.run/can-i-run/lfm2-24b-a2b/apple-m3-ultra-256gb)

Sources

Weights are measured from GGUF files; KV cache and overhead are computed, so totals can vary ~15% with context and runtime. Any tok/s is a bandwidth estimate. See methodology.