ISLAMABAD – Google has introduced its latest conversational artificial intelligence system, the Gemini 3.8 Live AI models, describing them as its most advanced live dialogue models to date.
The new models include Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, which are designed to make AI conversations more natural, faster and capable of handling complex tasks through voice interactions.
The launch follows Google’s earlier release of Gemini 3.1 Flash Live and expands the company’s focus on real-time AI assistants that can understand speech, process information and respond during ongoing conversations.
Gemini 3.8 Live Extended Thinking Brings Advanced AI Reasoning
The more advanced Gemini 3.8 Live Extended Thinking model is designed for tasks that require multiple steps and deeper reasoning.
Unlike traditional AI systems that complete reasoning before responding, Google says the model can think and communicate simultaneously.
This allows it to provide early responses such as acknowledging a request while continuing to work on complex tasks in the background.
The feature is aimed at improving experiences where users need AI assistance for longer workflows, including research, writing, document editing and productivity tasks.
Google Adds AI To Gmail, Docs And Keep
Google plans to use Gemini 3.8 Live Extended Thinking across several productivity features.
The model will support experiences such as:
- Gmail Live, which enables conversational email searches.
- Docs Live, designed for drafting and editing documents.
- Keep Live, which helps users create and organize notes.
The model is also beginning its rollout within Gemini Live, allowing users to interact with Google’s AI assistant through more advanced voice conversations.
Gemini 3.8 Live Improves Real-Time Conversations
The standard Gemini 3.8 Live AI models focus on speed, conversational ability and visual understanding.
Google says the model can process visual information in near real time while continuing background tasks, including tool usage and API calls.
This means users can ask questions, receive updates and continue conversations while the AI completes additional actions.
The model also supports automatic switching between 97 languages during the same conversation, improving accessibility for global users.
Gemini 3.8 Live Achieves Strong AI Benchmarks
Google said Gemini 3.8 Live Extended Thinking achieved strong results across several AI evaluation tests.
According to the company, the model scored:
- 82.6 on Artificial Analysis’ Speech-to-Speech Quality Index
- 68.6% on τ-Voice for agentic task completion
- 35.1% on Sierra’s τ-Voice banking benchmark
- 97.7% on Big Bench Audio reasoning evaluation
Google said these results demonstrate improvements in speech interaction, reasoning and AI task completion.
The company also highlighted the model’s competitive pricing compared with other advanced AI systems.
Gemini 3.8 Live Will Power Search AI Features
Google said Gemini 3.8 Live will also support the Search Live experience within AI Mode.
The integration aims to make search interactions more conversational by allowing users to ask questions naturally through voice while receiving AI-generated responses.
The move reflects Google’s broader strategy of integrating AI assistants across its search, productivity and communication products.
Google Expands Its AI Assistant Competition
The launch of Gemini 3.8 Live comes as major technology companies continue developing advanced voice-based AI systems.
Companies including OpenAI, Microsoft and Meta have also invested heavily in conversational AI models that can understand speech, process images and complete tasks.
With Gemini 3.8 Live, Google is focusing on making AI assistants more interactive by combining real-time conversation with deeper reasoning capabilities.
The Future Of Real-Time AI Conversations
The introduction of Gemini 3.8 Live AI models highlights the growing shift from text-based chatbots toward more natural AI assistants.
By combining voice interaction, visual understanding and advanced reasoning, Google aims to make AI tools more useful for everyday tasks.
As these models continue rolling out across Gemini and other Google products, they are expected to play a larger role in how users search, communicate and complete work online.
Key Points :
- Google launched Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking.
- The models are designed for more natural real-time AI conversations.
- Gemini 3.8 Live Extended Thinking can handle complex multi-step tasks.
- Google plans to integrate the technology into Gmail Live, Docs Live and Keep Live.
- Gemini 3.8 Live supports visual understanding and 97 languages.
- The models achieved strong results in speech and reasoning benchmarks.
- Gemini 3.8 Live will also power Search Live in AI Mode.
- Google is expanding AI assistants across its ecosystem.
Read More:
For more such exclusive articles, follow Upfront .
