Rust / Technical
LLM inference vs. the OOM killer
Mobile memory warnings and handling them in Rust.
Blog
Updates and technical notes from the NobodyWho team. Subscribe via RSS.
Comparison
NobodyWho vs Cactus: On-Device Inference Engine Comparison
NobodyWho vs Cactus compared on engine design, model format, hardware, platforms, cloud, and licensing.
Expo / Release
Announcing Expo support for NobodyWho
NobodyWho now works with Expo — run on-device LLMs in your Expo apps.
Feature / Release
Announcing Voice Activity Detection
Detect when someone starts and stops speaking, on-device.
Technical
Use fewer threads for CPU inference
How many worker threads should you use for CPU inference? Not all of them.
Feature / Release
Announcing Speech To Text & Text To Speech
STT & TTS in NobodyWho — easily generate and transcribe audio!
Apps / React Native
NobodyWho Chat app
NobodyWho Chat app is now available on mobile!
Apps / Swift
Apple Watch & Vision Pro apps
NobodyWho Swift bindings are now available! This article briefly introduces some of the interesting things we dealt with during development.
Kotlin / Release
Announcing Kotlin bindings for NobodyWho
NobodyWho now ships Kotlin bindings — run LLMs fully on-device in your Android and JVM apps, no cloud or server required.
Technical
LLM, give me a JSON. Make no mistakes.
So how exactly do you make your LLM output a JSON? What happens under the hood? And how do you make it reliable and fast? Diving into constrained sampling.