GPT-Live-1.1 adds new voices to ChatGPT Voice
Pick a new language-native voice in Settings > Voice. OpenAI's system card says GPT-Live-1.1 is the default for Plus and higher, and Quip for everyone else.
OpenAI is rolling out GPT-Live-1.1 and new language-native voices in ChatGPT Voice. To choose one, OpenAI says to "open Settings > Voice, then select Voice."
What changed#
On Oct 15 OpenAI's release notes said it is "rolling out GPT-Live-1.1 and GPT-Live-1.1 Quip, a mini version of the model, in ChatGPT Voice." At 2:19 PM ET, @OpenAI said GPT-Live-1.1 "brings these updates to Voice in ChatGPT, Codex, and dots" and is "Rolling out today to all users." The notes add that "Live also uses a wider range of short phrases to let you know when it's searching."
Before: GPT-Live-1 and GPT-Live-1 mini. After: the system card says "GPT-Live-1.1 is the default voice model for Plus and higher tier users" (including Codex and Dots), while "GPT-Live-1.1 Quip is the default voice model for others."
New Live voices cover French, German, Hindi, Japanese, Korean, and Mexican Spanish, plus Bubbie in English. OpenAI says "The voices you see may vary by region and account." A community moderator's announcement post adds that "Delegation is now more natural, and steerability has also improved."
Why it matters#
Our read: non-English accents and voice choice are the day-one change, and search status cues should repeat less. Per the system card, Plus and higher get 1.1 and everyone else, including Go, gets Quip. @OpenAI says the API is "Coming soon."
How to use it#
- Open Settings > Voice, then select Voice (per OpenAI's release notes).
- Pick a voice. The voices you see depend on your region and account.
- Our suggestion: start a Live call and ask something that needs search, to hear the new status phrases.
Prompt (ours, untested):
Speak with me in [language]. Use a natural [region] accent. When you need to search, keep status phrases short, then continue in that language. Confirm which Live model and voice you are using before we start.
Good to know#
- Who gets which model. Per the system card: 1.1 for Plus and higher (including Codex and Dots); Quip for others.
- API later. @OpenAI: "Coming soon to the API."
- Android quirk (user report). One forum user said Standard/Advanced Voice ignored the selected voice while Live Mode respected it. Not confirmed by OpenAI.
- Safety note (system card). On production voice-native evals the card says the models "perform at parity with their predecessors, with no statistically significant regressions." On synthetic evals it says "GPT-Live-1.1 has statistically significant regressions on sexual, mental health, emotional reliance, and gore categories compared to GPT-Live-1," and that OpenAI plans "to improve in these categories in future releases."
- Watch next: API for 1.1/Quip, and any fix for the Android voice-selection report.
Sources#
- https://help.openai.com/en/articles/6825453-chatgpt-release-notes (checked 2026-10-15 4:00 PM ET)
- https://openai.com/products/release-notes/ (checked 4:00 PM ET)
- https://x.com/OpenAI/status/2110797830361436604 (posted 2:19 PM ET / 18:19 UTC, checked 4:00 PM ET)
- https://community.openai.com/t/gpt-live-1-1-is-rolling-out-with-new-voices-and-better-accents/1406704 (created 3:11 PM ET, checked 4:00 PM ET)
- https://deploymentsafety.openai.com/gpt-live-1-1 (published Oct 15, 2026, checked 4:01 PM ET)
Last checked: 2026-10-15