OpenAI Adds 13 New Voices to Voice AI Models
The update improves natural language support while adjusting search tool availability for different user tiers.
Updated on Oct. 16, 2026 in Artificial Intelligence
Live Poll
Are you interested in using AI voice assistants that feature more natural human accents?
OpenAI has released the GPT-Live-1.1 and GPT-Live-1.1 Quip voice models, adding 13 new voices for use in ChatGPT and personal AI assistants. The update, which began rolling out globally on October 15, 2026, also introduces improved natural accents for several languages including French, German, and Japanese.
Why it matters
These models facilitate full-duplex, real-time voice interactions across a broader linguistic range, reflecting a push toward more seamless human-AI conversation. The release standardizes voice capabilities across different user tiers while adjusting the internal performance metrics for search integration.
The new GPT-Live-1.1 model serves as the default for Plus users, while the GPT-Live-1.1 Quip model serves other tiers. Tests indicate that GPT-Live-1.1 shows decreased performance in search tool availability versus its predecessor, whereas the Quip model shows improvements over the previous mini model.
The players
OpenAI
An AI research and deployment company that produces the ChatGPT platform and large language models for commercial and consumer use.
The details
Both models support full-duplex interaction, a communication mode where both parties can transmit and receive data simultaneously to allow for natural interruptions. Users access these capabilities by selecting the Voice option within settings, where the system triggers short auditory cues to signal when it is actively searching. Neither model currently supports video or screen-sharing functions.
Timeline
October 15, 2026: OpenAI began the global rollout of the GPT-Live-1.1 and GPT-Live-1.1 Quip models.
The Tech Race
This release signals an evolution in the GPT-Live voice model architecture, moving toward greater linguistic inclusivity. It follows a trajectory of iterative model updates aimed at refining the latency and naturalism of AI-driven voice interactions.
Users can access the 13 new voices immediately by navigating to the voice selection menu within their account settings. While these models support conversational English, French, German, Hindi, Japanese, Korean, and Mexican Spanish, they currently remain limited to voice input without video or screen-sharing capabilities.
The takeaway
The deployment of these models indicates that voice-first AI is shifting from a uniform standard toward tiered performance profiles for different user groups. Users should watch for upcoming documentation on API access, which OpenAI has confirmed as the next milestone for these voice architectures.
Further reading
For more on the current state of conversational agents, visit the Artificial Intelligence section.
Source note: This article includes information reported by Okay News.
Live Poll
Are you interested in using AI voice assistants that feature more natural human accents?
