text model · Nemotron 3 · iOS
Can I run Nemotron Cascade 2 30B-A3B on iPhone 15 Pro?
No. Nemotron Cascade 2 30B-A3B needs ~25.1 GB even at Q4_K_M, but iPhone 15 Pro only has ~4.5 GB usable.
Needs ~25.1 GB even at Q4_K_M, but only ~4.5 GB is usable.
That figure is at a 4k context and moves about ±15% as context length changes.
The gap is about 20.6 GB: Nemotron Cascade 2 30B-A3B needs roughly 25.1 GB at Q4_K_M and iPhone 15 Pro leaves only about 4.5 GB usable for a model. The lightest tracked hardware that runs Nemotron Cascade 2 30B-A3B is the Nvidia GeForce RTX 5090 (32GB) at 32 GB. See Nemotron Cascade 2 30B-A3B on Nvidia GeForce RTX 5090 (32GB).
- Q4_K_M needed
- ~25.1 GB
- Usable on device
- ~4.5 GB
- Device memory
- 8 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
- 30B (MoE, 3B active)
- Q4_K_M size
- 23.03 GB
- Q8_0 size
- 31.28 GB
- Context
- 256k
- Ollama tag
- nemotron-cascade-2:30b
- Memory
- 8 GB unified
- Usable for weights
- ~4.5 GB
- Power draw
- ~14 W
- Best runtime
- llama.cpp + Metal (via PocketPal or Off Grid app)
What you can run instead
Run Nemotron Cascade 2 30B-A3B on other hardware
FAQ
Can iPhone 15 Pro run Nemotron Cascade 2 30B-A3B?
No. Nemotron Cascade 2 30B-A3B needs ~25.1 GB even at Q4_K_M, but iPhone 15 Pro only has ~4.5 GB usable.
How much memory does Nemotron Cascade 2 30B-A3B need?
iPhone 15 Pro does not have enough memory. At Q4_K_M the weights are ~23.03 GB; with KV cache and runtime overhead, budget ~25.1 GB at a 4k context. It is a Mixture-of-Experts model (30B total / 3B active), so all experts must stay in memory; memory tracks total params, not active params.
What is the best tool to run Nemotron Cascade 2 30B-A3B 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/nemotron-cascade-2-30b-a3b/iphone-15-pro) Sources
- cpu-monkey.com
- developer.apple.com
- enclaveai.app
- forums.macrumors.com
- github.com/a-ghorbani
- github.com/google-ai-edge
- github.com/mainframecomputer
- gsmarena.com
- helloexpress.net
- huggingface.co/bartowski
- huggingface.co/nvidia
- layla-network.ai
- ollama.com
- privatellm.app
- support.apple.com
- techinsights.com
- wccftech.com
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.