Find out what’s new in the Gemini app in March's Gemini Drop.
Key Points
- Import chats & memories
- Personal Intelligence free in US
- Conversations 3.1 doubles context
Summary
March's Gemini Drop focuses on smoother, more personalized AI across Google apps and devices. Highlights for engineers include one-click transfer of external AI chat histories, wider Personal Intelligence integration (now free for U.S. users), enhanced Google TV visual/narration features, a new Lyria 3 Pro audio-generation capability, and faster, longer-context Conversations 3.1.
Key Points
- Transfer/import: Users can transfer AI memories and chat history from other providers in a few clicks — expect data-migration flows and consent UX.
- Personal Intelligence (U.S.): Now free for all Gemini users in the U.S.; connects across Gmail, Photos, and YouTube to provide context-aware assistance (requires cross-app access and user permissions).
- Google TV: Gemini-powered visual answers and narrated deep dives enable richer multimedia responses on TV platforms.
- Lyria 3 Pro: Audio-generation model now composes up to 3-minute tracks and can turn images or prompts into lyric-backed high-fidelity anthems (subscriber feature).
- Conversations 3.1: Faster response times and a context window that holds roughly twice as much history for more natural, continuous dialogue.
Engineering notes
- Expect increased auth/scope requirements for cross-app Personal Intelligence (Gmail/Photos/YouTube).
- Plan for larger session state and memory handling due to doubled context window; monitor latency and token usage.
- Multimedia features (TV visual answers, narrated deep dives, Lyria 3 Pro) will require pipelines for audio/video generation, delivery, and potentially DRM/asset storage.
- Check availability constraints (e.g., Personal Intelligence free in U.S., Lyria 3 Pro for subscribers) when gating features.
Where to start
- Review authentication and consent flows for cross-service data access.
- Audit session storage and rate limits to accommodate longer context retention.
- Prototype multimedia generation and delivery paths if integrating Gemini responses into apps or devices.