opnlocal

Run AI locally without learning how local AI works.

opnlocal checks your device, finds AI models that actually fit, downloads one, benchmarks it, and lets you chat privately on your own hardware.

Free and open source. No account. No telemetry.

All downloads, for every platform

Your hardware. Your model. Your AI.

There are hundreds of models and dozens of ways to shrink each one. opnlocal reads what your device has and offers only what fits, so you pick between a few plain choices instead of a spreadsheet.

opnlocal on a Windows PC: device check, model picks, download, speed test, then a chat
The real app on a Windows PC with a fresh install. The download and the speed test are sped up 30×; everything else is real time.

How it works

  1. Check your device

    opnlocal checks memory, graphics, processor and free storage.

  2. Tell it what you need

    Everyday help, coding, writing, or private documents.

  3. Pick from a few recommendations

    Not hundreds of model names and quantization options. Just Recommended, Lighter & faster, or Stronger, but tight.

  4. Chat locally

    The model runs on your own device. It's measured there first, so you know how fast it really replies.

Why opnlocal

Tools like Ollama and LM Studio are great when you already know which model and settings you want. opnlocal is for everyone else: it removes the technical choices instead of exposing them.

You decide

  • What you want AI for
  • Which of a few picks to try
  • What to ask

opnlocal handles

  • Choosing models that fit your hardware
  • VRAM, memory budgets and context sizes
  • Quantization and file formats
  • Downloading and verifying the file
  • Measuring real speed on your device
  • Running the chat locally, with no account

Your conversations stay on your device.

Read the complete list of network connections

Open source, Apache 2.0

The whole app, the engine and the model catalog are public. Read it, build it, or change it.

Support opnlocal

opnlocal is free and open source. If it saves you time or makes local AI easier to use, you can support development with a coffee.