$OPENAI
ChatGPT Voice Upgraded to GPT-Live-1.1, Adding New Voices Including Japanese and Korean
OpenAI announced on its official X account on October 15, U.S. time, the launch of its voice model GPT-Live-1.1, updating voice features in ChatGPT, Codex, and its always-on agent, dots. The new version has more natural-sounding accents and adds voices in French, German, Hindi, Mexican Spanish, Korean, and Japanese. It also says “checking…” less repetitively when searching. OpenAI says the update is rolling out to all users starting today, with the API version “coming soon.”
Free and Go users get mini Quip; Plus and above get GPT-Live-1.1
According to ChatGPT’s release notes, the update also introduces GPT-Live-1.1 Quip, a mini version. ChatGPT’s Live voice mode uses different versions depending on the user’s plan. Live will also use a greater variety of brief phrases when searching to let users know it’s looking something up.
OpenAI’s ChatGPT voice help page lists the usage limits for regular conversations on each plan, calculated over a rolling 24-hour period:
Plan | Voice model and usage
Free | GPT-Live-1.1 Quip, limited use; limits may be adjusted
Go | GPT-Live-1.1 Quip, 3 hours
Plus | GPT-Live-1.1, 3 hours
Pro ($100 per month) | GPT-Live-1.1, 15 hours
Pro ($200 per month) | GPT-Live-1.1, unlimited (subject to abuse-prevention safeguards)
For business plans, Business Standard and Business Premium include 3 hours and 15 hours, respectively. After that, 1.25 credits are deducted per minute. Enterprise, which is billed based on usage in U.S. dollars, costs $0.05 per minute.
New Japanese and Korean voices are available, and can be selected in Settings
According to the help page, the new voices added to Live mode include Aube and Rivage for French; Sonna and Miko for German; Leher and Neer for Hindi; Himari and Rindō for Japanese; Haetsal and Hanul for Korean; Tinta and Arrullo for Mexican Spanish; and an additional English voice, Bubbie. OpenAI says the voices shown to users may vary by region and account, and can be selected under “Settings → Voice.” No Chinese voices were added this time, but users can still ask ChatGPT to speak other languages during a conversation.
The help page also notes that Live mode currently does not support video or screen sharing. Mobile users who need these features can switch to the previous-generation Advanced Voice. ChatGPT’s default personality settings also do not currently apply to Live, but users can ask it to adjust its tone and speaking speed during a conversation.
GPT-Live-1 API access launched just last month
GPT-Live is OpenAI’s voice model that can listen and speak at the same time, respond in real time when interrupted by users, and delegate more difficult reasoning to a background model. OpenAI only made GPT-Live-1 available through its API in September, at $0.05 per minute for the front-end voice layer. This 1.1 version is launching first on ChatGPT, Codex, and dots; the API release date and pricing have not yet been announced.
This article, “ChatGPT Voice Upgraded to GPT-Live-1.1, Adding New Voices Including Japanese and Korean,” originally appeared on .
ChatGPT Voice Upgraded to GPT-Live-1.1, Adding New Voices Including Japanese and Korean
OpenAI announced on its official X account on October 15, U.S. time, the launch of its voice model GPT-Live-1.1, updating voice features in ChatGPT, Codex, and its always-on agent, dots. The new version has more natural-sounding accents and adds voices in French, German, Hindi, Mexican Spanish, Korean, and Japanese. It also says “checking…” less repetitively when searching. OpenAI says the update is rolling out to all users starting today, with the API version “coming soon.”
Free and Go users get mini Quip; Plus and above get GPT-Live-1.1
According to ChatGPT’s release notes, the update also introduces GPT-Live-1.1 Quip, a mini version. ChatGPT’s Live voice mode uses different versions depending on the user’s plan. Live will also use a greater variety of brief phrases when searching to let users know it’s looking something up.
OpenAI’s ChatGPT voice help page lists the usage limits for regular conversations on each plan, calculated over a rolling 24-hour period:
Plan | Voice model and usage
Free | GPT-Live-1.1 Quip, limited use; limits may be adjusted
Go | GPT-Live-1.1 Quip, 3 hours
Plus | GPT-Live-1.1, 3 hours
Pro ($100 per month) | GPT-Live-1.1, 15 hours
Pro ($200 per month) | GPT-Live-1.1, unlimited (subject to abuse-prevention safeguards)
For business plans, Business Standard and Business Premium include 3 hours and 15 hours, respectively. After that, 1.25 credits are deducted per minute. Enterprise, which is billed based on usage in U.S. dollars, costs $0.05 per minute.
New Japanese and Korean voices are available, and can be selected in Settings
According to the help page, the new voices added to Live mode include Aube and Rivage for French; Sonna and Miko for German; Leher and Neer for Hindi; Himari and Rindō for Japanese; Haetsal and Hanul for Korean; Tinta and Arrullo for Mexican Spanish; and an additional English voice, Bubbie. OpenAI says the voices shown to users may vary by region and account, and can be selected under “Settings → Voice.” No Chinese voices were added this time, but users can still ask ChatGPT to speak other languages during a conversation.
The help page also notes that Live mode currently does not support video or screen sharing. Mobile users who need these features can switch to the previous-generation Advanced Voice. ChatGPT’s default personality settings also do not currently apply to Live, but users can ask it to adjust its tone and speaking speed during a conversation.
GPT-Live-1 API access launched just last month
GPT-Live is OpenAI’s voice model that can listen and speak at the same time, respond in real time when interrupted by users, and delegate more difficult reasoning to a background model. OpenAI only made GPT-Live-1 available through its API in September, at $0.05 per minute for the front-end voice layer. This 1.1 version is launching first on ChatGPT, Codex, and dots; the API release date and pricing have not yet been announced.
This article, “ChatGPT Voice Upgraded to GPT-Live-1.1, Adding New Voices Including Japanese and Korean,” originally appeared on .