Google announced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, its latest conversational AI models focused on real-time voice interaction. Both models add upgraded reasoning, near real-time visual understanding, automatic language detection across 97 languages, and background tool calling that runs without interrupting the chat. A demonstration shows the Extended Thinking model acting as a programming tutor.
Google announced Gemini 3.8 Live with Live Avatar, pairing near real-time video generation with speech on its native live dialogue models. The feature creates animated personas with precise lip-syncing, natural expressions, and fluid turn-taking, and is available now in Gemini Enterprise. It supports 97 languages with seamless mid-conversation switching and asynchronous tool calling during dialogue.
Google announced Gemini 3.8 Live with Live Avatar, pairing near real-time video generation with speech on its native live dialogue models. The feature creates animated personas with precise lip-syncing, natural expressions, and fluid turn-taking, and is available now in Gemini Enterprise. It supports 97 languages with seamless mid-conversation switching and asynchronous tool calling during dialogue.
Medium post compares Gemini 3.8 Low and High operating modes and June–September 2026 updates
Medium post compares Gemini 3.8 Low and High operating modes and June–September 2026 updates
Google announced new productivity features for Gemini Live, its conversational AI assistant. The update aims to enhance user efficiency and task completion within the assistant.
Google announced new productivity features for Gemini Live, its conversational AI assistant. The update aims to enhance user efficiency and task completion within the assistant.
Google announced Gemini 3.5 Transcribe on August 26, 2026, a new speech-to-text model family that transcribes accurately in noisy environments, supports 85 languages, and auto-formats text by removing filler words. The models also handle complex details like phone numbers and custom vocabularies, and already power the Gboard 'Rambler' feature on Pixel 11.
Google announced Gemini 3.5 Transcribe on August 26, 2026, a new speech-to-text model family that transcribes accurately in noisy environments, supports 85 languages, and auto-formats text by removing filler words. The models also handle complex details like phone numbers and custom vocabularies, and already power the Gboard 'Rambler' feature on Pixel 11.
Alibaba's Qwen team announced Qwen3.8-Max, a new flagship model for coding and collaborative work, alongside the smaller Qwen3.8-27B. Both models are now live on Qwen Cloud, giving developers and researchers access to the latest Qwen family additions.
Alibaba's Qwen team announced Qwen3.8-Max, a new flagship model for coding and collaborative work, alongside the smaller Qwen3.8-27B. Both models are now live on Qwen Cloud, giving developers and researchers access to the latest Qwen family additions.