GenAI on Android with Gemini Nano and Firebase
summary
Compares on‑device Gemini Nano via ML Kit for four concrete APIs — summarize, proofread, rewrite, image description — noting it runs offline, currently supports 22 optimized devices, and requires model download before use; contrasts cloud using Gemini developer API or Vertex via Firebase AI Logic (three lines of Kotlin to call Gemini; live API streaming) which offers 1–2M token contexts, Imagen image generation with upcoming inpainting/outpainting, and a live demo pattern (four lines to wire actions) that works conceptually though the demo struggled with noise.
AI-written summary · about
chapters (11)
- 0:00GenAI on Android intro
- 1:02Why generative AI on mobile
- 2:27On-device vs cloud trade-offs
- 4:43MLKit GenAI API
- 9:42Gemini Nano devices
- 10:04Firebase AI Logic
- 14:18Live API demo
- 17:48Conversational AI in 4 lines
- 18:50Imagen model
- 21:21In-painting & out-painting
- 24:31Q&A
related talks
41mBuild intelligent Android apps with Google's AI
39mAdd Generative AI capabilities to your Android app
10mAI in your pocket: Building intelligent Android apps
41mGen AI for Android Developers
21mAdd Generative AI capabilities to your Android app
39mGetting Started with GenAI for Android Apps
35mBuilding Androidify: an AI-powered Android experience
18mLet's test Gemini, Google's AI in an Android application!