# OpenAI and Google upgrade AI voice tech to sound less like robots, more like coworkers > Source: > Published: 2026-08-10 14:27:00+00:00 Via openai.com # OpenAI and Google upgrade AI voice tech to sound less like robots, more like coworkers Both companies launched full-duplex voice models that listen and talk simultaneously, marking a shift from clunky turn-based AI conversations to something approaching actual dialogue. OpenAI launched GPT-Live on July 8, introducing GPT-Live-1 for paid subscribers and GPT-Live-1 mini for free users globally. Google, meanwhile, has been expanding its Gemini 3.5 Live Translate feature since June, adding multilingual real-time translation and deeper integration across its product ecosystem. ## What full-duplex actually means for users The technical term driving both announcements is “full-duplex architecture.” Older voice AI systems required users to speak, wait, then listen, a rigid back-and-forth that made conversations feel stilted. Full-duplex models can listen and speak at the same time, just like a human conversationalist who nods along, interjects with “right” or “got it,” and handles interruptions without derailing entirely. OpenAI’s GPT-Live models incorporate these acknowledgment cues and handle interruptions more gracefully than their predecessors. In human evaluations, the new models significantly outperformed previous voice systems on naturalness and conversational flow. One particularly clever feature: GPT-Live can delegate complex tasks mid-conversation. If a user asks something that requires a web search or deeper reasoning, the voice model hands the work off to a more capable model like GPT-5.5, then keeps talking while the answer is being fetched. Google’s approach leans heavily into multilingual capability. Gemini 3.5 Live Translate enables near-real-time speech translation across languages. The company has also pushed voice enhancements into practical tools like Docs Live for collaborative document editing and updated its home voice assistants with better context awareness. ## Safety features and the watermarking question OpenAI addressed provenance directly by adding SynthID watermarking technology to all GPT-Live-generated audio by July 31. SynthID embeds an imperceptible signal into AI-generated audio that allows downstream systems to detect its provenance. Both companies have also emphasized latency reductions in their new systems. Lower latency means shorter gaps between when a user speaks and when the AI responds, which is critical for making conversations feel natural. ## The competitive landscape heats up Both companies are targeting practical, everyday use cases rather than flashy demos. Language practice, hands-free assistance during commutes, complex workflow management — these are the scenarios being highlighted. For Google, the integration advantage is obvious. Gemini voice capabilities can be woven into Search, Docs, smart speakers, and Android devices. OpenAI’s strength lies in model quality and the developer ecosystem it has built around its API, which allows third-party applications to incorporate GPT-Live capabilities. OpenAI’s decision to offer GPT-Live-1 mini to free users worldwide is a strategic move worth noting. By making advanced voice capabilities available at no cost, OpenAI is betting that widespread adoption will create a flywheel effect. Google’s free tier for Gemini voice features follows similar logic. The improvements in context handling across both platforms also point toward a future where voice assistants maintain persistent understanding across sessions. Rather than treating each interaction as a blank slate, these systems are beginning to remember preferences, track ongoing tasks, and build on prior conversations. **Disclosure:** This article was edited by Editorial Team. For more information on how we create and review content, see our [Editorial Policy](https://cryptobriefing.com/editorial-policy/).