text model · DeepSeek-R1-Distill · iOS
Can I run DeepSeek-R1-Distill-Qwen 32B on iPhone 16?
No. DeepSeek-R1-Distill-Qwen 32B needs ~22.1 GB even at Q4_K_M, but iPhone 16 only has ~4.5 GB usable.
Needs ~22.1 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 17.6 GB: DeepSeek-R1-Distill-Qwen 32B needs roughly 22.1 GB at Q4_K_M and iPhone 16 leaves only about 4.5 GB usable for a model. The lightest tracked hardware that runs DeepSeek-R1-Distill-Qwen 32B is the Nvidia GeForce RTX 4090 (24GB) at 24 GB. See DeepSeek-R1-Distill-Qwen 32B on Nvidia GeForce RTX 4090 (24GB).
- Q4_K_M needed
- ~22.1 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
- 32B
- Q4_K_M size
- 19.85 GB
- Q8_0 size
- 34.82 GB
- Context
- 128k
- Ollama tag
- deepseek-r1:32b
- Memory
- 8 GB unified
- Usable for weights
- ~4.5 GB
- Power draw
- ~11 W
- Best runtime
- llama.cpp + Metal (via PocketPal or Off Grid app)
What you can run instead
Run DeepSeek-R1-Distill-Qwen 32B on other hardware
FAQ
Can iPhone 16 run DeepSeek-R1-Distill-Qwen 32B?
No. DeepSeek-R1-Distill-Qwen 32B needs ~22.1 GB even at Q4_K_M, but iPhone 16 only has ~4.5 GB usable.
How much memory does DeepSeek-R1-Distill-Qwen 32B need?
iPhone 16 does not have enough memory. At Q4_K_M the weights are ~19.85 GB; with KV cache and runtime overhead, budget ~22.1 GB at a 4k context.
What is the best tool to run DeepSeek-R1-Distill-Qwen 32B 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/deepseek-r1-distill-qwen-32b/iphone-16) 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.