OpenAI launches GPT-Live-1 API with full-duplex voice at $0.05 per minute
OpenAI released GPT-Live-1, a full-duplex voice model available via its API and platform, enabling simultaneous listening and speaking for real-time voice interactions. Priced at $0.05 per minute for the front-end voice layer, it supports custom voices, telephony integration, and delegation of complex reasoning to backend models like GPT-6 Astra. Yelp is using the feature for natural restaurant reservation calls.
Editorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page reads the event directly, while its address stays stable when the title changes.
- Summary covers the current reports
Cross-source coverage
Reporting timeline
OpenAI announces GPT-Live-1 voice feature for natural Yelp reservation calls
OpenAI's developer account announced a new AI model feature called GPT-Live-1, which enables real-time listening and interruption handling during voice conversations. The post highlights that Yelp is using this capability to make restaurant reservation calls feel more natural, allowing callers to interrupt, add details, or change direction mid-conversation. This development represents a practical application of advanced AI voice interaction technology in a commercial service context, specifically aimed at improving the user experience for making restaurant reservations through Yelp's platform.
Read sourceOpenAI releases GPT-Live-1 on API and platform with full-duplex voice capabilities
OpenAI has announced the availability of GPT-Live-1 on its API and OpenAI Platform. The new model brings ChatGPT's natural, full-duplex conversation capabilities to developers, allowing them to build applications and workflows powered by bidirectional voice mode. According to the announcement, GPT-Live-1 offers more control over how voice agents speak and act, enabling users to delegate tasks to other models. The release marks a significant expansion of OpenAI's voice AI offerings, moving beyond ChatGPT's native interface to allow third-party integration. The post highlights a cool user interface, suggesting enhanced developer experience. This development is expected to enable a new wave of voice-powered applications and automated workflows.
Read sourceGPT-Live-1 is now available in the API. Bring ChatGPT’s natural back-and-forth to your app, with voice agents that listen while they speak and work with the models and harness you choose. https://t.co/gIl1gwsBDV
OpenAI's developer account, OpenAIDevs, announced on X that GPT-Live-1 is now available in the API. The new model enables developers to integrate ChatGPT's natural conversational abilities into their applications, specifically supporting voice agents that can listen while speaking. The announcement highlights that the model works with the models and harness the developer chooses, offering flexibility. This release marks a significant update for developers building voice-enabled applications, allowing them to leverage OpenAI's advanced language model for more interactive and responsive user experiences. The post includes a link to further details, indicating a broader rollout of the technology.
Show 2 older updatesHide older updates
OpenAI Launches GPT-Live-1 Full-Duplex Voice Model in API at $0.05/Minute
OpenAI has announced the release of GPT-Live-1, a new full-duplex voice model available through its API. The model enables simultaneous listening and speaking, a significant advancement for real-time voice interactions. It supports delegating complex reasoning and tool calls to backend models such as GPT-6 Astra, allowing the front-end voice layer to focus on low-latency audio processing. The front-end voice layer is priced at $0.05 per minute. This release aims to simplify existing layered approaches for building voice agents by offering a single-model architecture. The announcement includes details on benchmark changes and pricing, providing developers with the information needed to evaluate the new model's capabilities and cost-effectiveness for their applications.
Read sourceOpenAI Announces GPT-Live-1 API with Full-Duplex Voice and Custom Voices
OpenAI has announced the release of GPT-Live-1, a new model available through its API that enables natural, full-duplex voice conversations. The model offers stronger instruction following compared to previous versions, along with support for custom voices and telephony integration. This development marks a significant step in making AI voice interactions more fluid and human-like, allowing developers to build applications that can handle real-time, two-way voice communication. The inclusion of telephony support suggests potential applications in call centers, virtual assistants, and other voice-based services. Custom voice capabilities allow for personalized user experiences. The announcement was made via OpenAI's official news channel, highlighting the company's continued focus on advancing conversational AI technologies for enterprise and developer use.
Read source