workflowstacks
design

omlx: Fast LLM Inference

Get fast LLM inference with omlx, ideal for founders working on Apple Silicon projects.
18,582 stars1,604 forksPythonGuide quality 8/10Updated 8/11/2026100% free · open source
What it does

omlx provides fast LLM inference for Apple Silicon projects, enabling continuous batching and SSD caching for efficient machine learning model deployment

When to use it
  • •When building Apple Silicon-based applications that require low-latency LLM inference
  • •When developing macOS or iOS apps that rely on large language models
  • •When optimizing machine learning workflows for Apple devices
Ready-to-paste prompt
python app.py --model gpt2 --batch-size 16 --ssd-cache /path/to/ssd
Heads up: Ensure you have Python 3.9 or later installed, as omlx relies on newer Python features for optimal performance
Saves to your device
Use with Claude
New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

🛠️ Technical setup

Expect 20–40 minutes in a terminal — or let your AI agent drive it.

Try it instantly — no install
Claude Code
mkdir -p ~/.claude/skills/omlx && curl -fsSL https://workflowstacks.com/api/skills/omlx/claude-skill -o ~/.claude/skills/omlx/SKILL.md
Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)
claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How omlx: Fast LLM Inference works
Codeflow
Free to inspect

Omlx: Fast LLM Inference is a very large Python project (~239k lines across 515 code files, plus 403 test files). It is a full software project: use it through its install path rather than reading it end to end. Last commit this month, Apache-2.0 license, has a test suite.

Size
Very large codebase
~239k lines · 515 code files · days to read — use, don't read
Setup
Developer setup
A real software project. Use it via its install path; don't expect to read it all.
Runs on
Python
No API keys detected
Python 77%Swift 8%HTML 7%C++ 3%JavaScript 3%
Where to start reading
  1. 1
    README.md
    Start here — what it does and how to install it
  2. 2
    omlx/cli.py
    Where the program starts running
  3. 3
    pyproject.toml
    Dependencies and the commands it exposes
  4. 4
    omlx/__init__.py
    Inside omlx/ — the main logic begins here
What's in each folder
omlx/Core code — the actual logic561 files
apps/Sub-packages / plugins152 files
docs/Documentation43 files
benchmarks/Evaluations & benchmarks8 files
scripts/Helper scripts5 files
packaging/Folder4 files
tools/Helper scripts2 files
Formula/Folder1 files
READMEHas testsDocumentedCI checksApache-2.0 licenseUpdated this month
Quick Actions
Details
Creator
jundot
Language
Python
Category
design
Published
2/13/2026

Are you the creator of this tool? Claim your listing → and earn 85% of every sale.

Related skills

More design tools founders pair with this one.

design★ 389,002
openclaw: Personal AI Assistant
Get a personal AI assistant for any OS and platform, ideal for founders seeking AI-driven support.
design★ 164,771
Stable Diffusion web UI for Design
Generate images with stable-diffusion-webui, a tool.
design★ 108,368
Supabase: Dedicated Postgres DB
Get a dedicated Postgres database with Supabase.
design★ 47,098
Summer2026-Internships: Find Top Talent
Get daily updated internship postings for software engineering and more.
design★ 39,133
Drawdb
Free, simple, and intuitive online database diagram editor and SQL generator.
design★ 22,019
Airbyte: Simplify Data Movement
Get seamless data integration for your startup with Airbyte.