Gemini Live gains memory access and expanded connected apps

At a glance:

• Gemini Live can now recall details from past chats for more personalized replies
• Expanded Connected Apps include YouTube, Utilities, Workspace and image generation
• Feature is live in English for US users, while Android settings still list memory as “coming soon”

What’s new for Gemini Live

Google has rolled out a major upgrade to its Gemini Live conversational AI, adding the ability to remember key details from earlier interactions. The change follows the Neural Expressive redesign and new voice options announced last month. In practice, Gemini Live can now reference prior chat history—such as dietary preferences, family birthdays, or favorite hobbies—so users no longer need to repeat information across sessions. Google’s support page notes that the assistant “follows the same permissions that you choose when you chat with Gemini,” meaning privacy controls remain consistent with the broader Gemini ecosystem.

Expanded Connected Apps integration

The update also broadens the range of Connected Apps Gemini Live can call upon during a conversation. The newly supported services are:

• YouTube
• Utilities
• Workspace (Docs, Sheets, Slides, etc.)
• Image generation tools
• Additional unnamed apps (Google indicates “more” is forthcoming)

While the integration now covers these platforms, Google Messages access is still pending. This expansion lets Gemini Live pull real‑time data—like a YouTube video link or a calendar entry—from the linked apps, enriching responses without the user having to switch contexts.

Availability and current limitations

The memory feature is currently live for English‑language users in the United States. In testing, the assistant successfully cited information from earlier chats, delivering “helpful, tailored responses.” However, the Personal Intelligence settings page on Android still labels Memory as “coming soon to Live,” suggesting that the functionality may not be fully enabled on all devices or may roll out gradually. Users should also note that the feature respects the same permission model as the standard Gemini chat, so any data access must be explicitly granted.

How this moves Gemini toward parity with the main chat

Gemini’s web‑based chat has long offered memory capabilities, allowing it to retain context across sessions. By bringing similar functionality to Gemini Live, Google narrows the gap between voice‑first and text‑first experiences. This parity is likely to encourage broader adoption of the live mode, especially for users who rely on voice interactions on mobile devices.

What to watch next

Google has hinted at further Connected Apps integrations and the eventual rollout of Memory on Android devices. Observers will be watching for:

1. Official support for Google Messages, which would close a major integration gap.
2. Expansion of the memory feature to additional languages and regions beyond the US.
3. Updates to privacy controls that clarify how remembered data is stored and protected.

Stakeholders—from developers building on Gemini’s APIs to end‑users seeking a more seamless AI assistant—should keep an eye on Google’s I/O announcements later this year for concrete timelines.