fastcontext-1.0-4b-sft-optiq-5bpw-mlx

microsoft/FastContext-1.0-4B-SFT, quantized with OptiQ by sahilchachra for Apple Silicon.

community

On disk2.9 GB

Memoryabout 5 GB

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 sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx

OptiQ Code

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

optiq serve --model sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx
optiq code

OptiQ Lab

optiq lab --model sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx

Claude Code

optiq run claude -m sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx

Codex

optiq run codex -m sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx

OpenCode

optiq run opencode -m sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx

Mistral Vibe

optiq run vibe -m sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx

Hermes Agent

optiq run hermes -m sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx

OpenClaw

optiq run openclaw -m sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx

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("sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx")
print(generate(model, tokenizer, prompt="Hello", max_tokens=256))

Details

Family	Qwen · Qwen docs
Architecture	Qwen3
Base model	microsoft/FastContext-1.0-4B-SFT
Size on disk	2.9 GB
Memory needed	about 5 GB
Images	no, text only
Downloads	6 in the last month
Updated	2026-06-16
Hugging Face	sahilchachra/fastcontext-1.0-4b-sft-optiq-5bpw-mlx