Skip to content

LFM2.5 24B A2B on Mac mini M4 24 GB: VRAM, speed and fit

LFM2.5 24B A2B at 4-bit runs on the M4 24 GB with 56K of context at about 56 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on M4 24 GB · 4-bit (Q4_K_M) · Ollama / LM Studio · Chat

GPU memory
23.2 GB
of 24.0 GB
Context
56K
auto
Speed
56 tok/s
33 when full
First reply
8.7 s
2K prompt

Where the memory goes

Model
13.5 GB
Context
1.1 GB
Overhead
570 MB
OS
8.0 GB
Free
835 MB
Memory
23.2 GB of 24.0 GB
Context
56K tokens · auto
Speed
56 tok/s · 33 when full
First reply
8.7 s for a 2K prompt
Running cost
$0.85/mo at home · $29/mo on a cloud RTX 2000 Ada

Validation: Not validated

What LFM2.5 24B A2B on Mac mini M4 24 GB was checked against
CheckTargetResultStatus
Fits in GPU memory16.0 GB usable15.2 GB neededPass
Headroom5% free835 MB free (5%)Pass
Context for chat16K or more56KPass
Writing speed15 tok/s or more56 tok/sPass
First replyunder 3 s, 2K prompt8.7 sFails

Targets: a reply you can read along with, a first reply the use case tolerates, and 5% of memory left free. Calculated from published specs, not benchmarked.

The setup

Model
LFM2.5 24B A2B
Parameters
24B, 2B active
Precision
4-bit (Q4_K_M)
Context
56K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
M4 24 GB
GPU can use
16.0 GB of 24.0 GB

Worth trying

  • 8-bit cache: Room for 104K context

Running cost

At home
$0.85/mo
39 W at the wall
Renting instead
$29/mo
RTX 2000 Ada at $0.24/h

4 h a day, power at $0.18 per kWh, cloud at RunPod on-demand prices as of Nov 22, 2026.

How to run it

Ollama
ollama run lfm2.5:24b
llama.cpp
llama-server -hf unsloth/LFM2.5-24B-A2B-GGUF
vLLM
vllm serve LiquidAI/LFM2.5-24B-A2B
MLX (Mac)
mlx_lm.chat --model mlx-community/LFM2.5-24B-A2B-OptiQ-4bit

LFM2.5 24B A2B on similar machines

Other models that run on Mac mini M4 24 GB

Questions

Can LFM2.5 24B A2B run on Mac mini M4 24 GB?

LFM2.5 24B A2B at 4-bit runs on the M4 24 GB with 56K of context at about 56 tokens a second.

Report VR-1Y946BV, calculated Nov 22, 2026 from published specs, not benchmarked.