Head of Realtime AI @OpenAI. Created WebRTC. Past: Distinguished Engineer @google, AIM. Amateur mathematician/musician.

Seattle, WA
Just posted our technical deep dive on the GPT-Live system, by @zahanm and yours truly.
GPT-Live can listen while it speaks. To make that feel natural at ChatGPT scale, we rebuilt the voice stack from client to model. This new architecture keeps audio flowing continuously, so deeper reasoning and tool use don't interrupt the conversation.
34
37
658
166,495
ChatGPT Voice can just do things! It now supports plugins, Work mode, and delegation to GPT-6 when more intelligence is needed.
We heard you loud and clear. ChatGPT Voice can now: - Use plugins like your email, calendar, and Slack. - Be powered by GPT-6 Astra, Sol, and Luna. - Be used in ChatGPT Work on web and mobile, so you can create docs, decks, sites, and spreadsheets or tackle complex tasks in the browser, just by talking. Rolling out globally today in the latest version of the app.
8
2
50
3,152
Justin Uberti retweeted
"stochastic parrot" was a mimetically-fit cognitive virus that spread from 2021-2025; it temporarily blinded many gifted people to the nature of AI progress, burning up crucial years in which they could have helped think through the response to the situation.
272
352
3,454
751,112
Justin Uberti retweeted
Little big thing: click your AirPods to mute voice ✨
New discovery today: Using ChatGPT Voice on iOS with AirPods, clicking an AirPod mutes you. This is great! Today ChatGPT helped me cook dinner. When I wanted to chat with my family, I muted via AirPod. Small thing, but now I can just leave Voice on and mute when needed.
12
3
148
18,753
Update: resolved
there's an issue with desktop voice when starting new sessions. Working on a mitigation right now...
1
2
28
4,418
there's an issue with desktop voice when starting new sessions. Working on a mitigation right now...
Replying to @keleftheriou
cc @juberti is there any known issue with Voice on Desktop right now?
1
2
14
7,505
Justin Uberti retweeted
you can use codex voice from your phone now, connecting to your computer from anywhere!! @axbehr and i got to star in this new codex ad, showing how much you can get done while working out. btw, this is powered by the new gpt-live-1
34
21
544
292,820
Custom API voices are already supported! developers.openai.com/api/do… (Restrictions apply, contact sales for more info)
Replying to @juberti
When are you going to buy @ElevenLabs so we can build our own custom voices? 😁
2
23
5,392
It's a whole new architecture that continuously generates output in real time: openai.com/index/continuous-… It shouldn't ever be slower than realtime though. The WebRTC and SIP endpoints will have the best performance.
Replying to @juberti
Any reason why GPT-Live API streams audio slower than realtime btw? Previously all OpenAI realtime models were faster than realtime...
3
35
5,416
We've added several new voices for the GPT-Live API - check them out in our new developer playground. platform.openai.com/audio/li…
23
15
129
17,177
There's an all-new playground for the GPT-Live API. You can pick your prompt and voice as usual, but you can also configure the delegation model and see the delegation interactions, as seen in this weather request. platform.openai.com/audio/li…
5
5
63
3,340
Now powered by the GPT-Live SIP (telephony) API!
📞 1-800-ChatGPT
3
3
84
14,251
yes, a watchdog timer is always a good idea. However you can probably just prompt GPT-Live to check if the user is still there and hang up the call if not.
Replying to @juberti
at $.05/minute, stuck sessions make silence detection a core cost control
1
8
1,795
This might be a bit quaint in the age of Codex, but here's a simple GPT-Live API demo of a voice agent that you can access via web or phone. web: hello-live.val.run phone: 425-800-0073 source: val.town/x/jubertioai/hello-… It's easy to customize - check out routes/agent.ts.
4
2
34
3,130
That's part of the magic - the realtime model can keep talking while delegation happens asynchronously in the background, and update seamlessly as new context streams in
Replying to @juberti
The $.05/min price is aggressive. Curious to see how the controllable delegation handles latency during model hand-offs in real-time voice.
4
1
33
3,629
The GPT-Live API is now... live! Our newest voice model offers better conversation, faster responses, and controllable delegation to any model. Best of all, it's also less expensive - only $.05/minute! openai.com/index/introducing…
GPT-Live-1 is now available in the API. Bring ChatGPT’s natural back-and-forth to your app, with voice agents that listen while they speak and work with the models and harness you choose.
27
18
195
44,927
There are a number of API changes for this model, and it needs to be prompted differently than the Realtime series. Check out our API overview doc to get started: developers.openai.com/api/do…
2
5
21
1,925
Firefox users sometimes failed to connect to ChatGPT Voice due to a 12-year-old WebRTC implementation issue. This prevented one of our team members from using his favorite browser, so he figured out what was going on and resolved this long-standing bug: bugzilla.mozilla.org/show_bu….
2
3
41
3,086
Just increased the ChatGPT Voice usage limits: Go: 3h Live mini Plus: 3h Live Pro 100: 15h Live Pro 200: Unlimited More details: help.openai.com/en/articles/…
Daily limits (as of July 13) —- Free: 2h Live mini Go/Plus: 1h Live, 2h Live mini Pro 200: unlimited Live (individual calls capped at 2h)
14
13
189
17,639