Skip to content
localmodel.run

audio model · whisper · iOS

Can I run Whisper large-v3-turbo on iPad Pro M4 (16GB, 1TB/2TB config)?

Compatibility verdict memory check
Yes, it runsheavy offloading

Yes. Whisper large-v3-turbo runs on iPad Pro M4 (16GB, 1TB/2TB config) at int8 (~1.5 GB of ~12 GB usable).

Needs ~1.5 GB Device usable ~12 GB

Runs at int8 using ~1.5 GB of ~12 GB usable.

Peak memory
~1.5 GB
Usable on device
~12 GB
Device memory
16 GB
Quant
int8

How to run it

Use whisper.cpp or faster-whisper at int8. It is light enough to run on CPU; a GPU just makes it faster.

Model whisper
Type
Speech to text
Parameters
809M
Peak memory
~1.5 GB at int8
License
MIT
Full Whisper large-v3-turbo requirements →
Device iOS
Memory
16 GB unified
Usable for weights
~12 GB
Best runtime
MLX (via Python or Swift; mlx-lm package)
Best models for iPad Pro M4 (16GB, 1TB/2TB config) →

You could also run

Run Whisper large-v3-turbo on other hardware

Apple M1 (8GB)8GB RAM Laptop (CPU/iGPU only)iPhone 15 ProiPhone 16iPhone 16 ProGeneric Android Phone (8GB RAM)

FAQ

Can iPad Pro M4 (16GB, 1TB/2TB config) run Whisper large-v3-turbo?

Yes. Whisper large-v3-turbo runs on iPad Pro M4 (16GB, 1TB/2TB config) at int8 (~1.5 GB of ~12 GB usable).

How much memory does Whisper large-v3-turbo need?

iPad Pro M4 (16GB, 1TB/2TB config) has room to spare. At int8 the realistic peak is ~1.5 GB of memory.

What do I use to run Whisper large-v3-turbo locally?

Whisper large-v3-turbo runs in whisper.cpp or faster-whisper (among others). It runs on CPU, so no GPU is required.

Sources

VRAM figures are sourced peak-usage anchors at the noted quant, validated 2026-06-15. See methodology.