The Pulse
OpenAI Rolls Out GPT-Live-1.1 With New ChatGPT Voice Options
OpenAI announced GPT-Live-1.1 and its smaller Quip version for ChatGPT Voice on October 15, 2026. The update adds voice options across six languages and English, though OpenAI’s release notes do not confirm a total of 13 new voices.
AI.info Team · Written with AI ·
We’re rolling out GPT‑Live‑1.1 and GPT‑Live‑1.1 Quip, a mini version of the model, in ChatGPT Voice, with more natural accents across languages.
OpenAI
OpenAI announced the rollout on October 15, adding two updated voice models and new voice options to ChatGPT. The company says the update also changes how Voice signals that it is searching, with a wider range of short phrases. The announcement describes more natural accents but does not publish benchmark results for accent quality or speech recognition.
GPT-Live-1.1 arrives in ChatGPT Voice
The release notes name GPT-Live-1.1 and GPT-Live-1.1 Quip, which OpenAI describes as a mini version of the model. The company presents both as part of the ChatGPT Voice rollout. It does not give a separate release date for either model or say in the announcement how long the rollout will take.
OpenAI’s system card places the update within a wider voice-model family. It describes GPT-Live models as full-duplex: they can listen and respond continuously, rather than waiting for a clearly defined turn to end. That design lets them handle pauses, interruptions and changes in speaking pace.
Six languages, plus an English voice
OpenAI lists new voice options for French, German, Hindi, Japanese, Korean and Mexican Spanish, as well as an English voice called Bubbie. The release notes say the voices available to an account may vary by region and account. Users can check the available choices under Settings, then Voice.
The announcement does not name the other new voices or state their total number. In particular, OpenAI’s release notes do not substantiate the claim that the update adds 13 voices. The company lists language options and Bubbie, but provides no complete voice inventory or count.
Quip serves as the smaller model
OpenAI’s system card says GPT-Live-1.1 succeeds GPT-Live-1, while GPT-Live-1.1 Quip succeeds GPT-Live-1 mini. The card identifies GPT-Live-1.1 as the default voice model for Plus and higher-tier users, including in Codex and Dots, and Quip as the default for other users. Those details describe model defaults; the release notes do not provide a plan-by-plan table of access to each new voice.
The difference matters because the update is not only a set of selectable accents. OpenAI is also replacing earlier models in its live voice lineup, with Quip occupying the smaller-model role. The company’s announcement does not disclose model size, response-time figures, pricing changes or a schedule for making the models available through its API.
OpenAI’s safety card reports mixed evaluation results
OpenAI says its production-prompt evaluations found the new models at parity with their predecessors, with no statistically significant regressions. On synthetic prompts designed to test difficult safety cases, however, the company reports statistically significant regressions for GPT-Live-1.1 in sexual content, mental health, emotional reliance and gore. OpenAI says those results need further research and warns that the synthetic tests do not predict performance across all production traffic.
The safety card adds a useful qualification to the product announcement: the new voices arrive alongside model changes whose performance varies across evaluation settings. OpenAI’s public release notes emphasize accents, search-status phrases and voice selection; they provide no audio samples, comparative listening tests or numerical quality measures. For now, the confirmed user-facing change is a new set of region-dependent voices in ChatGPT Voice, alongside GPT-Live-1.1 and its smaller Quip counterpart.