- Who it helps
- Developers designing voice interfaces and interactive assistants.
- Availability
- September 15 announcement, updated September 17, checked October 29. Integration, subscription eligibility and current prices were not tested.
Conversation and background tasks
Google separates a model aimed at efficient conversation from Extended Thinking for more complex tasks. The announcement describes visual context and background tool calls.
Confirm completed actions
Measure interruption recovery and whether completed actions match the request. A spoken progress update does not establish that a background action succeeded.
Gemini 3.8 Live: conversation while tools work ↗AI Tools Weekly September 14–20 ↗
What you can do
- Read the model access details before designing a voice integration. Open reference ↗
Where the evidence stops
Reported benchmarks were not reproduced. No voice session or API call was made.
Sources read. Sources checked October 29, 2026. No installation, model inference or end-to-end workflow was executed.
Sources & further reading
Web · official
Gemini 3.8 Live: conversation while tools work ↗
Establishes the subject, discovery context or stated verification limit.
blog.google · page checked 2026-10-29
More sources reviewed (1)
Web · analysis
AI Tools Weekly September 14–20 ↗
Establishes the subject, discovery context or stated verification limit.
Saaed Banbhan / Neodrop · page checked 2026-10-29
Coverage note: No qualifying current independent attention sample; original X posts unavailable. Reported outcomes were not independently reproduced.
The record over time (1)
Dated source checks and published explanations. An observation date can differ from the original release date.
- · Added to Current
Google’s Gemini 3.8 Live models are designed to keep a voice conversation going while background tasks run.