Step-3.7-Flash-OptiQ-3.7bpw-mlx

stepfun-ai/Step-3.7-Flash, quantized with OptiQ by lemuralabs for Apple Silicon; mixture-of-experts.

communityMoE

On disk106 GB

Memoryabout 124 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 lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-mlx

OptiQ Code

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

optiq serve --model lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-mlx
optiq code

OptiQ Lab

optiq lab --model lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-mlx

Claude Code

optiq run claude -m lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-mlx

Codex

optiq run codex -m lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-mlx

OpenCode

optiq run opencode -m lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-mlx

Mistral Vibe

optiq run vibe -m lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-mlx

Hermes Agent

optiq run hermes -m lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-mlx

OpenClaw

optiq run openclaw -m lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-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("lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-mlx")
print(generate(model, tokenizer, prompt="Hello", max_tokens=256))

Details

Family	Step3P7
Architecture	Step3P7
Base model	stepfun-ai/Step-3.7-Flash
Size on disk	106 GB
Memory needed	about 124 GB
Images	no, text only
Downloads	73 in the last month
Updated	2026-08-05
Hugging Face	lemuralabs/Step-3.7-Flash-OptiQ-3.7bpw-mlx