mlx-optiq
← Models

Mistral-Small-4-119B-2603-OptiQ-2bit

mistralai/Mistral-Small-4-119B-2603, quantized by OptiQ for Apple Silicon; takes images.

vision
On disk45.8 GB
Memoryabout 54 GB
Imagesyes

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/Mistral-Small-4-119B-2603-OptiQ-2bit

OptiQ Code

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

optiq serve --model mlx-community/Mistral-Small-4-119B-2603-OptiQ-2bit
optiq code

OptiQ Lab

optiq lab --model mlx-community/Mistral-Small-4-119B-2603-OptiQ-2bit

Claude Code

optiq run claude -m mlx-community/Mistral-Small-4-119B-2603-OptiQ-2bit

Codex

optiq run codex -m mlx-community/Mistral-Small-4-119B-2603-OptiQ-2bit

OpenCode

optiq run opencode -m mlx-community/Mistral-Small-4-119B-2603-OptiQ-2bit

Mistral Vibe

optiq run vibe -m mlx-community/Mistral-Small-4-119B-2603-OptiQ-2bit

Hermes Agent

optiq run hermes -m mlx-community/Mistral-Small-4-119B-2603-OptiQ-2bit

OpenClaw

optiq run openclaw -m mlx-community/Mistral-Small-4-119B-2603-OptiQ-2bit

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/Mistral-Small-4-119B-2603-OptiQ-2bit")
print(generate(model, tokenizer, prompt="Hello", max_tokens=256))

Details

FamilyMistral · Devstral · Mistral · Devstral docs
ArchitectureMistral 3
Base modelmistralai/Mistral-Small-4-119B-2603
Size on disk45.8 GB
Memory neededabout 54 GB
Imagesyes, through optiq serve and the Lab
Downloads1,386 in the last month
Updated2026-09-24
Hugging Facemlx-community/Mistral-Small-4-119B-2603-OptiQ-2bit