Run Gemma 4 on iPhone 17 with Google AI Edge Gallery

The standard iPhone 17 brings Apple's A19 chip and 8 GB of RAM to a non-Pro device for the first time. That combination handles Google Gemma 4 E2B effortlessly through Google AI Edge Gallery, with estimated speeds north of 35 tokens per second. The larger E4B model technically runs but eats most of your memory budget, so close other apps before loading it. For most users, E2B at full speed is the right pick.

Apple A198 GB RAM2025

Quick answer

On the iPhone 17 (Apple A19, 8GB RAM), run Google Gemma 4 E2B at ~35 tok/s (est.) via Google AI Edge Gallery. Gemma 4 E2B is ~2.5 GB on disk and uses 1-1.5 GB in memory, fully offline, no account required.

→ Install Google AI Edge Gallery (iOS 17+), then download Gemma 4.

TOP PICK

Gemma 4 E2B

SPEED

~35 tok/s (est.)

DEVICE RAM

8 GB

Speed is a ModelFit estimate from chip generation, not a measured benchmark.

Verdict

Recommended: Gemma 4 E2B

iPhone 17 with 8 GB RAM is the value sweet spot for on-device Gemma. E2B runs fast and cool; E4B works if you really need it.

Hardware Profile

DEVICE

iPhone 17

CHIP

Apple A19

RAM

8 GB

NEURAL ENGINE

16-core

Gemma 4 Performance on iPhone 17

Model	Download	RAM in Use	Speed	Source
Gemma 4 E2BTop Pick	~2.5 GB	1-1.5 GB	~35 tok/s	Estimated
Gemma 4 E4B	~5 GB	2-3 GB	~20 tok/s	Estimated

Speeds via Google AI Edge Gallery on iOS 17+. "Measured" numbers come from real-world Hacker News user reports; "Estimated" numbers are interpolations from chip generation. Both Gemma 4 variants use int4 quantization-aware training.

Best for

Daily-driver chat and writing assistance
Camera + AI: ask Gemma about anything you can photograph
Offline translation for travel
Voice memos with on-device transcription via Audio Scribe

Watch outs

E4B uses most of the 8 GB RAM. Close other apps first
Background apps may be evicted while a model is loaded
Wi-Fi recommended for the initial 2.5 GB E2B download

Setup Guide

Step-by-step install for Google AI Edge Gallery on iPhone 17, plus full benchmarks and the privacy details.

Read the full guide

Frequently Asked Questions

Does iPhone 17 support Google AI Edge Gallery?

Yes. iPhone 17 runs iOS 18+ and meets the Google AI Edge Gallery requirement of iOS 17.0 or later. Both Gemma 4 E2B and E4B variants are supported, though E2B is the recommended pick for the 8 GB RAM tier.

How fast is Gemma 4 E2B on iPhone 17?

Estimated at roughly 35 tokens per second based on the A19's improvements over the A18 (which measured ~30 tok/s on iPhone 16 Pro per Hacker News user reports). Real-world numbers will vary with thermal state and battery level.

Should I pick iPhone 17 or iPhone 17 Pro for AI?

iPhone 17 Pro with 12 GB RAM is the only model where Gemma 4 E4B runs comfortably. If you want the largest local model, get the Pro. If E2B is enough, and for most chat, translation, and image Q&A tasks it is, the standard iPhone 17 is the better value.

Can I run multiple AI models simultaneously on iPhone 17?

Not really. The Google AI Edge Gallery app loads one model at a time into memory. Switching between Gemma 4 E2B and E4B requires unloading and reloading, which takes a few seconds. With 8 GB RAM, holding two models in memory at once is not practical.