Apple Explores PrismML Technology to Run Larger AI Models Directly on iPhones

ArticleJoe RossignolNews

Apple has held discussions with startup PrismML to evaluate running substantially larger AI models directly on iPhones, according to a report by The Information. PrismML reportedly compressed Alibaba's 27-billion-parameter open-source model Qwen 3.6 to run entirely on an iPhone 17 Pro.

Key points

  1. PrismML compressed Alibaba's open-source LLM Qwen 3.6 (27 billion parameters) to run locally on an iPhone 17 Pro.
  2. Unlike Apple's on-device AFM 3 Core Advanced (a 20-billion-parameter sparse model activating 1 to 4 billion parameters at a time), PrismML runs all 27 billion parameters simultaneously.
  3. Apple's AFM 3 Core Advanced powers iOS 27 capabilities, including expressive Siri AI voices and systemwide dictation on iPhone 17 Pro and iPhone Air models.
  4. Deploying larger on-device models allows more Apple Intelligence features to bypass Private Cloud Compute, reducing Apple's cloud costs and increasing user privacy.

Why it matters

Expanding on-device processing limits reliance on cloud infrastructure like Apple's Private Cloud Compute, cutting operational expenses while keeping user data securely on the device.

Analysis

Mobile on-device LLMs have typically relied on sparse architectures to operate within hardware constraints. Running a dense model with 27 billion active parameters on a smartphone demonstrates significant compression progress, bridging the performance gap between edge devices and cloud-hosted AI.

Mentioned links

Found in

Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.

Related entries

Similar in content, found by the AI search.

  • Link:PocketAiHub

    Qwen3.8-27B-Abliterated-MLX: Uncensored Multimodal Model for Apple Silicon

    PocketAiHub has released an unofficial MLX derivative family of Qwen3.8-27B with suppressed refusal behaviour ('abliteration'). The consolidated repository provides quantized builds ranging from experimental 2-bit AWQ to full BF16 optimised for Apple Silicon.

    AI & LLMs· Resource

  • X post:Hamid Maei

    Ternary Bonsai 2 27B: Highly compressed Qwen3.8 model reduced to 5.9 GB

    PrismML has announced Ternary Bonsai 2 27B, a compressed version of Qwen3.8 27B. It is nine times smaller than the full-precision original, requiring only 5.9 GB of storage while retaining 98.2% of aggregate benchmark performance.

    97bookmarks21kviews

    AI & LLMs· Announcement

  • Link:drumih

    TurboFieldfare: Running Gemma 4 26B in 2 GB RAM on Apple Silicon

    TurboFieldfare is an open-source inference engine written in Swift and Metal that executes 4-bit Gemma 4 26B-A4B-IT on Apple M-series Macs using roughly 2 GB of RAM. While the model weights require 14 GB of disk space, the engine keeps only the shared model base and KV cache in memory and streams routed expert layers on demand directly from the SSD.

    AI & LLMs· Tool

  • Repository:drumih/turbo-fieldfare

    TurboFieldfare: Gemma 4 26B-A4B Inference in 2 GB RAM on Apple Silicon

    TurboFieldfare is a dedicated Swift and Metal runtime designed to run Google's Gemma 4 26B-A4B on Apple Silicon Macs within an approximately 2 GB RAM budget. It avoids loading the full 14.3 GB model into memory by keeping only the 1.35 GB core and KV cache in RAM while streaming the required mixture-of-experts weights from SSD on demand.

    6.6kstarsSwift

    AI & LLMs· Tool

Back to the knowledge base

Let's talk about your project

Get in touch

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.