Home · Support · Privacy Policy

Need help with onLM? Find answers below or reach out to us directly.

Contact Support

For any questions, issues, or feedback, please email us at [email protected]. We typically respond within 24–48 hours.

Frequently Asked Questions

How does onLM work?

onLM runs AI language models directly on your iPhone or iPad. Once you download a model, everything happens on-device — no internet connection required for conversations. Your data never leaves your phone.

Which devices are supported?

onLM requires iOS 26.0 or later. It works on all iPhones and iPads, but performance varies by device. Devices with 8 GB RAM or more provide the best experience with a wider selection of models.

How much storage do I need?

Model sizes range from ~1.5 GB to ~7 GB. We recommend having at least 5 GB of free storage before downloading your first model. You can delete models anytime to free up space.

Which model should I choose?

onLM automatically recommends the best model for your device based on available memory. As a general guide:

• 4 GB RAM (e.g. older iPhones): Qwen 2B or Phi 4 Mini — fast and lightweight
• 8 GB RAM (e.g. iPhone 16 Pro): Qwen 4B — excellent all-around choice
• 12+ GB RAM (e.g. iPad Pro M2+): Qwen 9B or Gemma 12B — most capable models

What is Apple Intelligence in onLM?

On supported devices running iOS 26+, onLM can use Apple's built-in Foundation Models. This option requires no download and works instantly. It appears automatically when your device supports Apple Intelligence.