{"slug": "google-unveils-gemini-3-8-live-and-3-8-live-thinking-ai-models", "title": "Google unveils Gemini 3.8 Live and 3.8 Live Thinking AI models", "summary": "Google launched the Gemini 3.8 Live and Gemini 3.8 Live Thinking AI models, adding real-time audio transcription, reasoning and translation plus multi-step task handling, the company said on 17 September 2026. Google reported an average Word Error Rate of 4.0% for streaming transcription and 2.6% for non-streaming transcription, with support for 85+ languages and automatic detection of 97 languages mid-conversation. All audio generated by the new models is watermarked with SynthID to flag AI-generated content and curb misinformation.", "body_md": "## Quick summary - click for full details\n\nConcise summary of key highlights\n\nIn one line\n\nGoogle launches Gemini 3.8 Live and 3.8 Live Thinking AI models for real-time audio reasoning and multi-step tasks.\n\nKey points\n\n• Gemini 3\n\nEnables real-time audio transcriptions, reasoning, and translations for multi-utility apps with near-instant visual input processing.\n\n• Gemini 3\n\nDesigned for high-complexity tasks with advanced multi-step reasoning and increased intelligence.\n\n• Gemini 3\n\nDedicated speech-to-text model offering highly precise transcription across 85+ languages with low error rates.\n\n• Voice command integration\n\nUsers can invoke Gemini AI via smartphones to perform complex multi-step tasks across Google and third-party apps using voice commands.\n\n• Watermarking for AI content\n\nAll audio generated by the new models is watermarked with SynthID to detect AI-generated content and prevent misinformation.\n\nKey statistics\n\n4.0%\n\nAverage Word Error Rate (WER) for streaming transcription\n\n2.6%\n\nAverage Word Error Rate (WER) for non-streaming transcription\n\n85+\n\nNumber of languages supported by Gemini 3.5 Transcribe\n\n97\n\nNumber of languages automatically detected mid-conversation\n\nProcessed with AI. Reviewed by DH Digital Team.\n\nPublished 17 September 2026, 05:28 IST", "url": "https://wpnews.pro/news/google-unveils-gemini-3-8-live-and-3-8-live-thinking-ai-models", "canonical_source": "https://www.deccanherald.com/technology/artificial-intelligence/google-unveils-gemini-38-live-and-38-live-thinking-ai-models-4149118", "published_at": "2026-09-17 05:28:19+00:00", "updated_at": "2026-09-17 05:53:25.087184+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "generative-ai", "ai-products", "natural-language-processing"], "entities": ["Google", "Gemini 3.8 Live", "Gemini 3.8 Live Thinking", "Gemini 3.5 Transcribe", "SynthID"], "alternates": {"html": "https://wpnews.pro/news/google-unveils-gemini-3-8-live-and-3-8-live-thinking-ai-models", "markdown": "https://wpnews.pro/news/google-unveils-gemini-3-8-live-and-3-8-live-thinking-ai-models.md", "text": "https://wpnews.pro/news/google-unveils-gemini-3-8-live-and-3-8-live-thinking-ai-models.txt", "jsonld": "https://wpnews.pro/news/google-unveils-gemini-3-8-live-and-3-8-live-thinking-ai-models.jsonld"}}