LFM2.5-24B-A2B-OptiQ-4bit

LiquidAI/LFM2.5-24B-A2B, quantized by OptiQ for Apple Silicon; mixture-of-experts.

MoE

On disk16.7 GB

Memoryabout 21 GB

Capability Score74.59

Capability Score breakdown

The six metrics behind this model's Capability Score of 74.59, as reported in the model card. Values are compared with uniform 4-bit.

This model
Uniform 4-bit

Use it

Install

pip install mlx-optiq

Serve it

An OpenAI- and Anthropic-compatible server on this Mac. Every app below can use it.

optiq serve --model mlx-community/LFM2.5-24B-A2B-OptiQ-4bit

OptiQ Code

The terminal coding agent uses whatever model optiq serve is running.

optiq serve --model mlx-community/LFM2.5-24B-A2B-OptiQ-4bit
optiq code

OptiQ Lab

optiq lab --model mlx-community/LFM2.5-24B-A2B-OptiQ-4bit

Claude Code

optiq run claude -m mlx-community/LFM2.5-24B-A2B-OptiQ-4bit

Codex

optiq run codex -m mlx-community/LFM2.5-24B-A2B-OptiQ-4bit

OpenCode

optiq run opencode -m mlx-community/LFM2.5-24B-A2B-OptiQ-4bit

Mistral Vibe

optiq run vibe -m mlx-community/LFM2.5-24B-A2B-OptiQ-4bit

Hermes Agent

optiq run hermes -m mlx-community/LFM2.5-24B-A2B-OptiQ-4bit

OpenClaw

optiq run openclaw -m mlx-community/LFM2.5-24B-A2B-OptiQ-4bit

optiq run starts optiq serve if it is not running, sets the agent up for this session only, and stops the server when the agent exits.

mlx-lm, text only

from mlx_lm import load, generate

model, tokenizer = load("mlx-community/LFM2.5-24B-A2B-OptiQ-4bit")
print(generate(model, tokenizer, prompt="Hello", max_tokens=256))

Details

Family	LFM2.5 · LFM2.5 docs
Architecture	LFM2 MoE
Base model	LiquidAI/LFM2.5-24B-A2B
Size on disk	16.7 GB
Memory needed	about 21 GB
Images	no, text only
Capability Score	74.59 (from the model card)
Downloads	452 in the last month
Updated	2026-10-16
Hugging Face	mlx-community/LFM2.5-24B-A2B-OptiQ-4bit