Skip to main content
webAI runs AI models locally on your device. The right model depends on your hardware — specifically how much memory you have and which inference backend you’re using. This guide helps you pick the best option.

​
Quick recommendation

Not sure where to start? Here’s the short version:
Your deviceRecommended modelBackend
MacBook Air (8 GB)Gemma 3n E2B or Qwen3 4BMLX
MacBook Pro (16 GB)Qwen3 8BMLX
MacBook Pro (32 GB+)Qwen3 32BMLX
Any Mac (fallback)Qwen3 4Bllama.cpp
Browser onlyQwen3 1.7B or Qwen2.5 0.5B
When in doubt, start with a smaller model. You can always switch to a larger one later — and smaller models load faster and respond more quickly.

​
Downloading a model

1

Open Settings

Navigate to Settings from the launcher or sidebar.
2

Go to the AI section

Find the model management panel where you can see available and downloaded models.
3

Select a model

Choose a model from the list. The UI shows the model size, memory requirement, and compatible backends.
4

Download

Click download. The model is saved to your device — once downloaded, it’s available offline.

​
Understanding model sizes

Larger models are generally smarter but require more memory and respond more slowly. Here’s what to expect:
Size classParametersGood for
Small (0.5B - 1.7B)Fast, lightweightQuick answers, simple tasks, low-memory devices
Medium (4B - 8B)BalancedGeneral use, research, writing, code
Large (12B - 32B)More capableComplex reasoning, detailed analysis, nuanced writing
Very large (70B+)Most capableAdvanced reasoning, demanding tasks (requires 64+ GB)

​
Choosing a backend

The backend determines how the model runs on your hardware. See On-Device AI for the full technical breakdown.
BackendPlatformSpeedModel support
MLXmacOS (Apple Silicon)FastestUp to 235B parameters
llama.cppmacOS (desktop app)ModerateUp to 70B parameters
WebGPUAny browserVaries by GPUUp to 1.7B parameters
The system selects the best backend automatically based on your hardware. You can override this in Settings if you prefer a specific backend.

​
Managing your model library

​
Switching models

You can switch between downloaded models at any time from the Oasis settings or the model selector. The current model unloads and the new one loads in its place.

​
Storage

Models are stored locally on your device. A small model (0.5B) takes around 500 MB of disk space, while a large model (32B) can take 20+ GB. Check your available disk space before downloading larger models.

​
Removing models

If you need to free up space, you can delete downloaded models from the model management panel. They can always be re-downloaded later.

​
LoRA adapters

After choosing a base model, you can optionally attach a adapter through your persona configuration. Adapters specialize the model for specific tasks without requiring a separate download of a full model.

​
Learn more

On-Device AI

Technical deep dive into backends, tiers, and hardware routing.

Oasis

The app where you interact with your local AI.