Google launches Live voice AI in Gmail, Docs and Keep and releases Gemini 3.8 Flash
Google brought Live voice AI to Gmail, Docs and Keep, and released Gemini 3.8 Flash, calling it its best reasoning and coding model yet.
Gmail Live answers natural-language questions about a person's inbox. In Engadget's description, a user can ask when their child's next school event is and get a spoken summary based on relevant email, with links to the messages shown in a live transcript. The conversation can continue with follow-up questions, and the assistant can be interrupted to change subjects. Docs Live can create a first draft or restructure a document from conversational instructions and, if given permission, draws on Gmail, Drive, Chat and the web. Keep Live records spoken notes and can infer a grocery list from a request such as adding ingredients for pesto.
Gmail Live is available to Google AI Plus, Pro and Ultra subscribers. Docs Live and Keep Live are available to Google AI Pro and Ultra subscribers. Engadget said the features have come out of beta for paying subscribers, and Google said they will also reach Workspace Business customers soon.
Availability details differ between reports. TechCrunch said all Live features would be available in English on iOS and Android. The Verge and Engadget said Docs Live is available on both platforms but that Keep Live is Android-only for now.
TechCrunch noted that Google has been adding voice features across its apps, including cross-app dictation for its Mac app in April, a Pixel 11 dictation tool called Rambler last month, and the Gemini 3.5 Transcribe model for speech-to-text in August.
Also on Sept. 3, according to Slashdot, Google released Gemini 3.8 Flash, described as its third Flash model in six weeks. It comes in a standard workhorse edition for tasks such as agentic operations and software development, plus a Cyber edition tuned for vulnerability detection and mitigation. Slashdot said Google called the model its best reasoning and coding model yet and claimed it tops the DeepSWE leaderboard.
Developer API access is priced at an introductory rate of $0.75 per million input tokens and $3.75 per million output tokens through the end of the year, according to the Slashdot report, with regular pricing of $1.50 and $7.50 after that. Gemini 3.8 Flash is available across Google's ecosystem, while the Cyber variant is limited to trusted testers and governments. Access in the Gemini app requires a Pro or Ultra subscription, and free experimentation remains available in AI Studio.