Chrome Silently Downloads 4GB Gemini Nano AI Model
Google Chrome is quietly downloading a 4GB Gemini Nano model to qualifying devices without user consent, and re-download…
🏷️ DEV
978 articles about 'DEV'
Memento Builds a Shared Brain for Your AI Tools
Open-source project Memento offers active memory collection and cross-IDE sharing, challenging passive memory tools like…
Snapdragon X Elite Delivers 45 TOPS AI to Laptops
Qualcomm's Snapdragon X Elite chip brings 45 TOPS of dedicated AI processing to Windows laptops, reshaping the on-device…
OpenAI Codex CLI Gains Autonomous Multi-Step Engineering
OpenAI upgrades Codex CLI with autonomous multi-step software engineering capabilities, enabling end-to-end coding workf…
Apple Intelligence Gives Siri On-Device App Control
Apple expands Siri with on-device AI capabilities that let the assistant control third-party apps without cloud processi…
Cursor AI Hits 10M Developers, Reshaping Coding
Cursor AI code editor crosses the 10 million active user milestone, signaling a major shift in how developers write soft…
Where Developers Get AI News in 2025
A growing wave of developers are building AI news aggregators, highlighting the challenge of keeping up with the fast-mo…
GitHub Copilot Workspace Goes Fully Autonomous
GitHub Copilot Workspace now enables developers to manage entire code repositories autonomously, from planning to deploy…
Microsoft Phi-5 Beats GPT-4 on Edge Devices
Microsoft's new Phi-5 small language model delivers GPT-4-level performance while running locally on smartphones and lap…
Hyundai Brings On-Device AI Assistant to Cars Globally
Hyundai Motor Group integrates an on-device AI assistant into its connected vehicle platform, enabling offline functiona…
Lumi API Launches With 4M Free Token Offer
New AI API platform Lumi debuts with 2M free tokens on signup and 2M per referral, targeting developers seeking affordab…
Does Cursing at AI Make Its Code Worse?
Developers report that rude prompts lead to buggier AI-generated code, sparking debate about how tone affects LLM output…