On-device AI Models for Android and Kotlin
Kotlin libraries that run models on device across Android and the JVM through ONNX Runtime.
The Kotlin SDKs run on Android and the JVM through ONNX Runtime, published via JitPack with the `ai.desertant` package namespace. Bundled models keep inference fully on device, with no network call at runtime.
Add the dependency and call it from Kotlin or Java. The same models and behavior you get on Apple and the web, native to Android.
Available models
Reversible PII redaction across 24 EU languages, from a LiteRT model bundled for Android and the JVM.
Emoji suggestions from text in 23 languages, on device via ONNX Runtime.
Content topics from any text across 101 languages on LiteRT, with a ~15 MB English-only build for English apps.
Single-stroke shape recognition for Android canvases and Compose.
Speech enhancement on Android through ONNX Runtime.
Language identification for short text across 84 languages, from a 2 MB model bundled in the AAR.
Pricing
Every model is free up to 100k monthly active devices per SDK. Unlimited inference per user.