text model · North · iOS
Can I run North Mini Code 1.0 on iPhone Air?
No. North Mini Code 1.0 needs ~21.1 GB even at Q4_K_M, but iPhone Air only has ~8 GB usable.
Needs ~21.1 GB even at Q4_K_M, but only ~8 GB is usable.
That figure is at a 4k context and moves about ±15% as context length changes.
The gap is about 13.1 GB: North Mini Code 1.0 needs roughly 21.1 GB at Q4_K_M and iPhone Air leaves only about 8 GB usable for a model. The lightest tracked hardware that runs North Mini Code 1.0 is the Nvidia GeForce RTX 4090 (24GB) at 24 GB. See North Mini Code 1.0 on Nvidia GeForce RTX 4090 (24GB).
- Q4_K_M needed
- ~21.1 GB
- Usable on device
- ~8 GB
- Device memory
- 12 GB
Which quant fits
How to run it
On iOS use Apple Foundation Models (Built into iOS 26, ~3B on-device model, zero download, fully private.).
- Parameters
- 30.5B (MoE, 3B active)
- Q4_K_M size
- 19 GB
- Q8_0 size
- 32 GB
- Context
- 500k
- Ollama tag
- north-mini-code-1.0
- Memory
- 12 GB unified
- Usable for weights
- ~8 GB
- Best runtime
- llama.cpp + Metal (via PocketPal or Off Grid app)
What you can run instead
Run North Mini Code 1.0 on other hardware
FAQ
Can iPhone Air run North Mini Code 1.0?
No. North Mini Code 1.0 needs ~21.1 GB even at Q4_K_M, but iPhone Air only has ~8 GB usable.
How much memory does North Mini Code 1.0 need?
iPhone Air does not have enough memory. At Q4_K_M the weights are ~19 GB; with KV cache and runtime overhead, budget ~21.1 GB at a 4k context. It is a Mixture-of-Experts model (30.5B total / 3B active), so all experts must stay in memory; memory tracks total params, not active params.
What is the best tool to run North Mini Code 1.0 on iOS?
On iPhone and iPad, Apple Foundation Models (Built into iOS 26, ~3B on-device model, zero download, fully private.) is the standard choice. Phones realistically run 1B-4B class models. Anything larger thermally throttles or OOMs.
Embed this
A live badge for your model card or README, updated as the data is. [](https://localmodel.run/can-i-run/north-mini-code-1.0/iphone-air) Sources
Weights are measured from GGUF files; KV cache and overhead are computed, so totals can vary ~15% with context and runtime. Any tok/s is a bandwidth estimate. See methodology.