Run an LLM Locally on Android and iPhone (August 2026)
Apple caps its on-device model at 4096 tokens per session, and Google's own tables put Gemma 4 E2B at 25.0 decode tokens/sec on an iPhone 17 Pro CPU.
Tag
Apple caps its on-device model at 4096 tokens per session, and Google's own tables put Gemma 4 E2B at 25.0 decode tokens/sec on an iPhone 17 Pro CPU.
Apple's new Extensions system lets rival AI chatbots integrate with Siri. The access is real, but so are the limitations that keep Apple's assistant in control.
Apple released iOS 26.4 on March 24 but the Gemini-powered Siri overhaul is nowhere to be found. Features now pushed to iOS 26.5 or later.
From Chat & Ask AI's 300 million exposed messages to widespread hardcoded secrets, security researchers reveal a systemic failure across AI applications
Apple partnered with Google to rebuild Siri using Gemini AI, but the March release is missing key features while privacy questions remain unanswered.
Xcode 26.3 introduces agentic coding, letting AI agents build projects, run tests, search docs, and iterate on fixes autonomously through the open Model Context Protocol.