Startups Race to Ship On-Device AI Features
A growing number of startups are moving artificial-intelligence features off the cloud and onto the device itself, betting that users will prefer tools that work without sending personal data to a remote server.
The shift is partly technical and partly about trust. Smaller, more efficient models now run acceptably on modern phones and laptops, cutting latency and removing the recurring cost of cloud inference. For founders, that changes the economics of building an AI product.
Challenges remain — on-device models still trail the largest cloud systems on the hardest tasks — but for everyday features like summarising, transcribing and search, several teams argue local is now good enough, and increasingly expected.
