Yhn TechnologyMobile first seen 3 d ago, last 23 min ago, peak #8

Qwen 27B model said to run on iPhone 18 Neural Engine

Original: Qwen3.8-27B on the iPhone 18 Neural Engine at 15–70 tok/s (1.5 bpw)

A claim circulating among AI enthusiasts says a Qwen 3.8 model with 27 billion parameters can run locally on the iPhone 18's Neural Engine, generating text at 15 to 70 tokens per second using extreme 1.5-bit quantization. If accurate, it would mark a notable step for on-device large language models, though independent benchmarks have not yet confirmed the figures.

Why now: unclear

QwenAppleiPhone 18

Open on hn →

Rank over time, top of the chart is #1. 62 snapshots from 3 d ago to 23 min ago.

Evidence

• Qwen3.8-27B on the iPhone 18 Neural Engine at 15–70 tok/s (1.5 bpw) · anemll · 30

API: https://socialmediatrends-api.osmike.com/v1/trends/2642601