mlx-optiq
Pre-built quants · Hugging Face

OptiQ quants, ready to load.

Every model is a standard MLX checkpoint. Load it with mlx_lm.load(...), no special runtime and no PyTorch. OptiQ measures each layer's sensitivity and spends bits where they matter, so a mixed-precision quant keeps the accuracy a flat uniform quant gives up, at the same size.

Load any quant · with optiq
import optiq  # unlocks every OptiQ arch
from mlx_lm import load, generate

model, tok = load(
  "mlx-community/Laguna-XS-2.1-OptiQ-4bit")
print(generate(model, tok, prompt="Hi"))
$ optiq convert Qwen/Qwen3.6-27B --target-bpw 4.5
12+
model families
2–8
bit, mixed precision
Model families

Pick a family. Details in the docs.

Each family page lists every quant, its size, and its six-metric Capability Score against stock uniform 4-bit. Here is the lineup, newest first.

01

Muse-Glimmer

A 30B image-text reasoning model, Capability Score 87.36. Sliding-window attention on three layers in four keeps memory flat as context grows.

30B · vision · read the docs →
02

Nemotron

NVIDIA's Mamba-attention hybrid. A dense 4B, a 30B-A3B MoE, and the 3.5 Lightning 30B-A3B, each with a KV config for its few attention layers.

4B + two 30B MoE · read the docs →
03

LFM2.5

Liquid AI's on-device series, 230M through 8B, plus a 3B that takes images. Convolutional blocks alternate with full attention, so only a few blocks hold a KV cache and it stays small as context grows.

230M to 8B · vision · on-device · read the docs →
04

Qwen

Three generations in one family: 3.8 with image input, 3.6, and 3.5, the daily driver behind Conjure and most of our agent work. 0.8B through 122B.

3.5 · 3.6 · 3.8 · dense + MoE · vision · read the docs →
05

Gemma-4

Google's instruct and QAT series, e2b through 31B, with image input on the multimodal quants.

e2b to 31B · vision · read the docs →
06

DeepSeek-V4

A 43-layer sparse-attention MoE that runs three different attention modes depending on the layer. The routed experts carry almost all the weight, so they go to 2-bit while attention stays higher.

Flash · 2-bit experts · read the docs →
07

Laguna

poolside's Laguna-XS-2.1, a sparse-MoE reasoning model for coding and agents. Its OptiQ quant nearly ties our best Capability Score.

MoE · coding · read the docs →
08

Diffusion LLM

Our first non-autoregressive family. A discrete-diffusion Gemma that denoises a canvas of tokens instead of decoding left to right.

26B-A4B · non-autoregressive · read the docs →
09

Mistral · Devstral

Devstral-Small-2, the tool-calling coding model. It was the first OptiQ quant of a Mistral-family base.

24B · agentic coding · read the docs →
10

MiniCPM5

A strong sub-gigabyte 1B with a hybrid <think> reasoning mode. The smallest full-capability quant in the lineup.

1B · on-device · read the docs →
11

Nanbeige

A looped transformer. It runs a 22-layer stack twice with shared weights, so it computes like a deeper network without the parameters.

3B · looped · read the docs →
12

Mage-VL

Microsoft's 5B vision-language model, the first video model in the lineup. It reads images and video, with the vision tower kept at bf16 in a sidecar.

5B · video + image · read the docs →

And dozens more across the ecosystem, from us and from developers publishing their own quants. Browse every OptiQ quant on Hugging Face →

Load

Standard MLX. Run mlx_lm.load("mlx-community/<repo>") and generate. Vendored-arch families (Laguna, Nanbeige, dhara) need import optiq first to register the arch.

Serve

Point optiq serve at any repo for an OpenAI and Anthropic-compatible endpoint with a mixed-precision KV cache.

Tune

Sensitivity-aware LoRA, hot-swappable adapters, and a full local Lab, all on your Mac.