# \#gpt-live

**URL:** https://community.openai.com/tag/gpt-live/1198.md

[Latest](https://community.openai.com/latest.md) · [Categories](https://community.openai.com/categories.md) · [Tags](https://community.openai.com/tags.md)

---

## [Eunice is almost here (William Gibson, Agency)](https://community.openai.com/t/eunice-is-almost-here-william-gibson-agency/1409651)

<div class="topic-metadata">

**Author:** [@platypus](https://community.openai.com/u/platypus)\
**Replies:** 1\
**Last updated:** [October 25, 2026, 8:58pm UTC](https://community.openai.com/t/eunice-is-almost-here-william-gibson-agency/1409651 "2026-10-25T20:58:42Z")

</div>

I was reading William Gibson’s 2020 novel “Agency” (sequel/prequel to The Peripheral - also VERY good) and came across this passage between Verity (the human) and Eunice (AI assistant), and thought we are actually living…

---

## [GPT-Live-1.1 is rolling out with new voices and better accents](https://community.openai.com/t/gpt-live-1-1-is-rolling-out-with-new-voices-and-better-accents/1406704)

<div class="topic-metadata">

**Author:** [@VeitB](https://community.openai.com/u/VeitB)\
**Replies:** 32\
**Last updated:** [October 22, 2026, 11:12pm UTC](https://community.openai.com/t/gpt-live-1-1-is-rolling-out-with-new-voices-and-better-accents/1406704 "2026-10-22T23:12:31Z")

</div>

New voices. Better accents. And less “checking…” GPT-Live-1.1 is an update for ChatGPT, Codex, and dots, introducing 13 new voices and considerably better performance in French, German, Hindi, Mexican Spanish, Korean, J…

---

## [GPT-Live starts calls talking about a buffet in a strange voice?](https://community.openai.com/t/gpt-live-starts-calls-talking-about-a-buffet-in-a-strange-voice/1402575)

<div class="topic-metadata">

**Author:** [@JakeTheSnake](https://community.openai.com/u/JakeTheSnake)\
**Replies:** 11\
**Last updated:** [October 22, 2026, 12:38pm UTC](https://community.openai.com/t/gpt-live-starts-calls-talking-about-a-buffet-in-a-strange-voice/1402575 "2026-10-22T12:38:53Z")

</div>

Ive had 2/251 GPT-Live calls (Appliance Repair CSR Ops) that start with a massive artifact where the voice is super exasperated talking about a ‘flexible buffet’, and its weird it happened twice because the script/phrasi…

---

## [Feature request: GPT-Live "delegate-only" mode (model hears and reports, speaks only appended commentary)](https://community.openai.com/t/feature-request-gpt-live-delegate-only-mode-model-hears-and-reports-speaks-only-appended-commentary/1408207)

<div class="topic-metadata">

**Author:** [@skyzer](https://community.openai.com/u/skyzer)\
**Replies:** 0\
**Last updated:** [October 20, 2026, 9:43pm UTC](https://community.openai.com/t/feature-request-gpt-live-delegate-only-mode-model-hears-and-reports-speaks-only-appended-commentary/1408207 "2026-10-20T21:43:32Z")

</div>

Category: API. Title: GPT-Live: a session mode where the model never speaks on its own and says only session.commentary.append content (delegation: client) Use case. We run phone workflows on GPT-Live-1 through the Live…

---

## [Live/sessions with Responses Delegation does not support MCP tool type](https://community.openai.com/t/live-sessions-with-responses-delegation-does-not-support-mcp-tool-type/1408038)

<div class="topic-metadata">

**Author:** [@MoulayIdris](https://community.openai.com/u/MoulayIdris)\
**Replies:** 0\
**Last updated:** [October 20, 2026, 11:47am UTC](https://community.openai.com/t/live-sessions-with-responses-delegation-does-not-support-mcp-tool-type/1408038 "2026-10-20T11:47:59Z")

</div>

Hello, Apologies if this has already been raised. Docs and (not so much) the API reference for GPT-live-1 suggest we should be able to use MCP servers in Responses Delegation config however both the python sdk types (as…

---

## [Tool call parameters with descriptions in gpt-live](https://community.openai.com/t/tool-call-parameters-with-descriptions-in-gpt-live/1402940)

<div class="topic-metadata">

**Author:** [@melindacr](https://community.openai.com/u/melindacr)\
**Replies:** 2\
**Last updated:** [October 19, 2026, 11:18pm UTC](https://community.openai.com/t/tool-call-parameters-with-descriptions-in-gpt-live/1402940 "2026-10-19T23:18:14Z")

</div>

Greetings. I am working on migrating a phone agent for customer intake and appointment bookings from gpt-realtime-1.5 to gpt-live, and I’m wondering how folks are handling lists of tool call parameters with gpt-live. In…

---

## [How can browser echo be fixed in GPT Live calls?](https://community.openai.com/t/how-can-browser-echo-be-fixed-in-gpt-live-calls/1403507)

<div class="topic-metadata">

**Author:** [@abstamal](https://community.openai.com/u/abstamal)\
**Replies:** 8\
**Last updated:** [October 15, 2026, 2:43pm UTC](https://community.openai.com/t/how-can-browser-echo-be-fixed-in-gpt-live-calls/1403507 "2026-10-15T14:43:57Z")

</div>

Been building a phone receptionist for clinics on GPT-Live for the last few weeks. The voice part is really good. Full duplex makes it feel like a real person on the line, and people talk over it without it falling apart…

---

## [Gpt-live-1 responses delegation: voice model sometimes never delegates (acknowledges verbally, sends nothing)](https://community.openai.com/t/gpt-live-1-responses-delegation-voice-model-sometimes-never-delegates-acknowledges-verbally-sends-nothing/1398585)

<div class="topic-metadata">

**Author:** [@matteo89](https://community.openai.com/u/matteo89)\
**Replies:** 6\
**Last updated:** [October 15, 2026, 1:56pm UTC](https://community.openai.com/t/gpt-live-1-responses-delegation-voice-model-sometimes-never-delegates-acknowledges-verbally-sends-nothing/1398585 "2026-10-15T13:56:36Z")

</div>

We’re running gpt-live-1 over WebSocket (/v1/live/sessions, audio/pcmu @ 8kHz, telephony) with responses-mode delegation (gpt-5.6 backend, tool\_choice: auto, tools for transfer/hangup/knowledge lookups). Intermittently …

---

## [GPT-Live greeting delayed and often glitched (SIP)](https://community.openai.com/t/gpt-live-greeting-delayed-and-often-glitched-sip/1398920)

<div class="topic-metadata">

**Author:** [@slam\_9e](https://community.openai.com/u/slam_9e)\
**Replies:** 2\
**Last updated:** [October 14, 2026, 1:25pm UTC](https://community.openai.com/t/gpt-live-greeting-delayed-and-often-glitched-sip/1398920 "2026-10-14T13:25:19Z")

</div>

We’re connecting to gpt-live over SIP. When asking gpt-live to speak first, we’re encountering two issues: There’s always a 2–3s silence, then the greeting starts at 00:03 of the call. In about half the calls, the gree…

---

## [\[VIDEO\] GPT-Live product Ball 2 on Shark Tank Premiere](https://community.openai.com/t/video-gpt-live-product-ball-2-on-shark-tank-premiere/1403863)

<div class="topic-metadata">

**Author:** [@klangdon79](https://community.openai.com/u/klangdon79)\
**Replies:** 0\
**Last updated:** [October 6, 2026, 7:36pm UTC](https://community.openai.com/t/video-gpt-live-product-ball-2-on-shark-tank-premiere/1403863 "2026-10-06T19:36:58Z")

</div>

Ball 2 is a sensor packed foam ball that connects with a mobile app. The mobile app allows anyone to describe a physics based game via GPT-Live conversation. The game requirements get coded with other LLM …

---

## [Introducing GPT-Live-1 in the API](https://community.openai.com/t/introducing-gpt-live-1-in-the-api/1396471)

<div class="topic-metadata">

**Author:** [@sps](https://community.openai.com/u/sps)\
**Replies:** 16\
**Last updated:** [October 6, 2026, 6:57pm UTC](https://community.openai.com/t/introducing-gpt-live-1-in-the-api/1396471 "2026-10-06T18:57:45Z")

</div>

GPT-Live-1 brings ChatGPT’s natural, full-duplex voice conversations to the API. It listens and speaks simultaneously, handles pauses, interruptions, and backchannels, and adapts when a conversation changes direction. …

---

## [Feature Request: Multi-Participant Live Voice — Speaker-Aware Group Conversations](https://community.openai.com/t/feature-request-multi-participant-live-voice-speaker-aware-group-conversations/1396988)

<div class="topic-metadata">

**Author:** [@yu.xie](https://community.openai.com/u/yu.xie)\
**Replies:** 1\
**Last updated:** [October 6, 2026, 5:41pm UTC](https://community.openai.com/t/feature-request-multi-participant-live-voice-speaker-aware-group-conversations/1396988 "2026-10-06T17:41:16Z")

</div>

Summary I would like to propose a new capability for ChatGPT Live Voice: Multi-participant conversations in which ChatGPT can distinguish known speakers and participate naturally as one member of the group. This builds…

---

## [GPT-Live-1 still struggles with customer names and digit sequences — what works?](https://community.openai.com/t/gpt-live-1-still-struggles-with-customer-names-and-digit-sequences-what-works/1402189)

<div class="topic-metadata">

**Author:** [@raoul](https://community.openai.com/u/raoul)\
**Replies:** 1\
**Last updated:** [October 1, 2026, 2:15pm UTC](https://community.openai.com/t/gpt-live-1-still-struggles-with-customer-names-and-digit-sequences-what-works/1402189 "2026-10-01T14:15:39Z")

</div>

I use GPT-Live-1 in ChatGPT Voice. When I say a person’s name or spell it aloud, Live often gets the spelling wrong. Sequences of digits are also unreliable for me. ChatGPT’s regular voice dictation is almost flawless in…

---

## [iOS Voice → Live: strong turn-taking/micro-naturalness, but weak macro conversational state tracking (“mode drift”)](https://community.openai.com/t/ios-voice-live-strong-turn-taking-micro-naturalness-but-weak-macro-conversational-state-tracking-mode-drift/1398844)

<div class="topic-metadata">

**Author:** [@rikkarta](https://community.openai.com/u/rikkarta)\
**Replies:** 2\
**Last updated:** [September 28, 2026, 10:55am UTC](https://community.openai.com/t/ios-voice-live-strong-turn-taking-micro-naturalness-but-weak-macro-conversational-state-tracking-mode-drift/1398844 "2026-09-28T10:55:56Z")

</div>

Platform / version: iOS, ChatGPT app 2026.244 (33940143573) Voice mode: Voice → Live (no voice change mid-call) Session length: approximately 30–60 minutes Context / limit notice: none shown Memory: enabled globally …

---

## [Viola feels more alive — please bring this ChatGPT voice to Live Mode](https://community.openai.com/t/viola-feels-more-alive-please-bring-this-chatgpt-voice-to-live-mode/1401113)

<div class="topic-metadata">

**Author:** [@Merpherus](https://community.openai.com/u/Merpherus)\
**Replies:** 0\
**Last updated:** [September 26, 2026, 10:54pm UTC](https://community.openai.com/t/viola-feels-more-alive-please-bring-this-chatgpt-voice-to-live-mode/1401113 "2026-09-26T22:54:19Z")

</div>

I wanted to share some positive feedback about Viola, a voice I can use in ChatGPT in Brazil. It sounds fantastic to me. Viola feels more alive. The pacing and intonation sound natural, with a spontaneous, curious, play…

---

## [GPT-Live-1 “responses” delegation returns 403 for a model that works fine direct](https://community.openai.com/t/gpt-live-1-responses-delegation-returns-403-for-a-model-that-works-fine-direct/1399936)

<div class="topic-metadata">

**Author:** [@federicodepro](https://community.openai.com/u/federicodepro)\
**Replies:** 2\
**Last updated:** [September 22, 2026, 8:50pm UTC](https://community.openai.com/t/gpt-live-1-responses-delegation-returns-403-for-a-model-that-works-fine-direct/1399936 "2026-09-22T20:50:12Z")

</div>

\# Bug report: GPT-Live-1 “responses” delegation returns 403 for a model that works fine directly ## Summary \`POST /v1/live/sessions\` with \`delegation.type: “responses”\` accepts the session creation request (\`201\`) for …

---

## [Gpt-live-1 switches between masculine and feminine voices mid-call and repeats speech](https://community.openai.com/t/gpt-live-1-switches-between-masculine-and-feminine-voices-mid-call-and-repeats-speech/1399151)

<div class="topic-metadata">

**Author:** [@robert.lukoshko](https://community.openai.com/u/robert.lukoshko)\
**Replies:** 0\
**Last updated:** [September 19, 2026, 5:59pm UTC](https://community.openai.com/t/gpt-live-1-switches-between-masculine-and-feminine-voices-mid-call-and-repeats-speech/1399151 "2026-09-19T17:59:27Z")

</div>

We are using gpt-live-1 for live voice conversations and are hearing an intermittent, very noticeable voice-consistency problem. During a single call, the assistant unexpectedly switches from a masculine-sounding voice …

---

## [GPT-Live-1: best voice and prompting for natural French](https://community.openai.com/t/gpt-live-1-best-voice-and-prompting-for-natural-french/1397735)

<div class="topic-metadata">

**Author:** [@ezzoelyas](https://community.openai.com/u/ezzoelyas)\
**Replies:** 3\
**Last updated:** [September 19, 2026, 10:19am UTC](https://community.openai.com/t/gpt-live-1-best-voice-and-prompting-for-natural-french/1397735 "2026-09-19T10:19:44Z")

</div>

Hi, I’m building a French voice agent with GPT-Live-1 through the API. My main issue is pronunciation and accent. I’m trying to get natural standard French, with a professional and polished tone, but even with detailed…

---

## [Support for end of turn in gpt-live voice model](https://community.openai.com/t/support-for-end-of-turn-in-gpt-live-voice-model/1397759)

<div class="topic-metadata">

**Author:** [@anatRvz](https://community.openai.com/u/anatRvz)\
**Replies:** 0\
**Last updated:** [September 15, 2026, 9:47am UTC](https://community.openai.com/t/support-for-end-of-turn-in-gpt-live-voice-model/1397759 "2026-09-15T09:47:21Z")

</div>

Hey there, I’m migrating my product from realtime to live model and one thing I’m really missing is the ability to know when a turn from the voice model is over. More specifically, I’m having a very hard time understan…

---

## [Brain context compaction in gpt-live-1](https://community.openai.com/t/brain-context-compaction-in-gpt-live-1/1396632)

<div class="topic-metadata">

**Author:** [@anatRvz](https://community.openai.com/u/anatRvz)\
**Replies:** 3\
**Last updated:** [September 14, 2026, 4:42pm UTC](https://community.openai.com/t/brain-context-compaction-in-gpt-live-1/1396632 "2026-09-14T16:42:25Z")

</div>

Hey there, We are moving forward on migrating from gpt-realtime to gpt-live and one thing that seems missing in the docs and the new API is how to manage the context the brain behind the voice model has. If I understan…

---

## [Feature Request: Multi-Speaker Identity Profiles for ChatGPT Live Voice](https://community.openai.com/t/feature-request-multi-speaker-identity-profiles-for-chatgpt-live-voice/1396987)

<div class="topic-metadata">

**Author:** [@yu.xie](https://community.openai.com/u/yu.xie)\
**Replies:** 0\
**Last updated:** [September 12, 2026, 3:08pm UTC](https://community.openai.com/t/feature-request-multi-speaker-identity-profiles-for-chatgpt-live-voice/1396987 "2026-09-12T15:08:16Z")

</div>

Summary I would like to propose an extension to the idea of Primary Speaker Recognition: Allow ChatGPT Live Voice to recognize multiple explicitly enrolled speakers, rather than only distinguishing the primary user from…

---

## [GPT Live struggles to follow even a simple daily workflow](https://community.openai.com/t/gpt-live-struggles-to-follow-even-a-simple-daily-workflow/1395674)

<div class="topic-metadata">

**Author:** [@xief](https://community.openai.com/u/xief)\
**Replies:** 0\
**Last updated:** [September 8, 2026, 4:50am UTC](https://community.openai.com/t/gpt-live-struggles-to-follow-even-a-simple-daily-workflow/1395674 "2026-09-08T04:50:00Z")

</div>

I’ve been trying to use GPT Live for a simple daily planning routine, but the biggest problem is the lack of reliable, persistent memory. It keeps forgetting the workflow I’ve established, making it difficult to use for …
