A new AI model named Maple-Preview, a 20 billion parameter Mixture of Experts (MoE) model, has been demonstrated running efficiently on an iPhone. The model achieved a speed of 120 tokens per second, showcasing its capability for on-device AI processing. AI
IMPACT Demonstrates the increasing feasibility of running sophisticated AI models directly on mobile devices.
RANK_REASON Demonstration of an AI model running on consumer hardware, not a core AI release from a frontier lab.
Read on Mastodon — fosstodon.org →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
