Models & Labs
Apple Releases 20B On-Device Model in Just 2 GB
Lev Selectorhigh confidence
Why it matters
- →Enables advanced AI features on existing hardware without upgrades
- →Enhances user privacy by keeping data on-device
- →Sets a new standard for model compression and efficiency
Apple has showcased a significant advancement in on-device AI efficiency by releasing a 20-billion parameter model that operates within a mere 2 GB of memory footprint. This breakthrough enables complex AI tasks to run directly on consumer hardware without cloud dependency, enhancing privacy and reducing latency. The development is critical for expanding the capabilities of personal devices like iPhones and Macs.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1Apple's AI Chips Emerge from Self-Driving Car Effort
The Verge AI · July 12, 2026 · Related
- 2Apple Introduces M6 and M5 Ultra Chips
Matt Wolfe · August 28, 2026 · Background
- 3Apple's AI-Focused Mac Minis
The AI Daily Brief · August 29, 2026 · Background
- 4OpenAI buys Macs for private AI agents
Wes Roth · September 1, 2026 · Background
- 5Bonsai 2 and Needle: Tiny Local AI Models
Lev Selector · September 25, 2026 · Background
- 6Tony Fadell critiques failed AI hardware and on-device future
TechCrunch AI · October 7, 2026 · Background
- 7Liquid AI releases efficient on-device agent model
Hugging Face Blog · October 13, 2026 · Background
- 8Meta and Apple collaborate on running Muse AI on Macs
The Verge AI · October 15, 2026 · Background
More from Lev Selector
© Lev SelectorInvestment
Market & Regulationbusiness
OpenAI Raises Capital at $1.25 Trillion Valuation
OpenAI has completed a new funding round valuing the company at $1.25 trillion, reflecting massive investor demand.
Lev Selector
© Lev SelectorModels & Labsmodels
Google Gemini 4 Carbon Leads LM Arena Leaderboard
Google's new model, Gemini 4 Carbon, has taken the top spot on the LMSYS Chatbot Arena leaderboard.
Lev Selector
© Lev SelectorOpen Sourcemodels
Mistral Large 4 Released as Open Weights Model
French AI startup Mistral has released its latest large language model, Large 4, with open weights available for download.
Lev Selector
More in Models & Labs
Models & Labsmodels
llama.cpp b12029 adds CUDA 13 and Snapdragon support
This release quietly cements llama.cpp as the universal inference runtime by finally supporting NVIDIA's latest CUDA 13 stack alongside AMD's ROCm 10.0. The inclusion of native builds for Linux arm64 with Snapdragon chips, covering CPU, Adreno GPU, and Hexagon NPU, signals a serious push into mobile edge deployment. While Apple Silicon KleidiAI is disabled in this specific build, the broader hardware coverage means developers no longer need to compile from source for most modern accelerators. It’s less about new features and more about removing friction for anyone running local models on bleeding-edge or diverse hardware.
llama.cpp Releases
© The Verge AIModels & Labsother
Apple Intelligence arrives in China via iOS 27.3
Apple is finally unlocking its AI suite for Chinese users with the iOS 27.3 beta, marking a significant regulatory milestone after months of approval delays. The rollout relies on Alibaba’s Qwen model rather than OpenAI, integrating features like notification summaries and visual intelligence directly into the OS. By sidestepping the new Siri in favor of established text and image tools, Apple is prioritizing immediate utility over the full generative experience. This move effectively brings Apple’s AI capabilities to a massive new market, albeit with localized constraints that distinguish it from the global release.
The Verge AI
© Matt WolfeModels & Labsmodels
Google Releases Gemini 4 Carbon for Cost-Efficient Inference
Google launched Gemini 4 Carbon, a new model variant aimed at providing high performance at a lower cost for inference tasks.
Matt Wolfe