Standard Voice retirement will break workflows for Pro users

Community
, , , ,
1

Standard Voice is being retired on Sept 9. For many of us who rely on voice-first use, this is a big problem:

Standard Voice gave accurate GPT-5 output I could trust.

GPT Voice for me often feels unhelpfully stylized, and lacks the completeness, accuracy and integrity of the raw model.

Custom GPT actions don’t work in GPT Voice, which breaks entire workflows.

Transcription in Standard Voice has recently degraded, making even that unreliable.

I’m asking OpenAI to:

  1. Offer a Plain Mode in GPT Voice that gives unstyled, raw GPT-5 output.

  2. Enable actions in GPT Voice, so Pro workflows don’t collapse.

  3. Maintain accurate speech-to-text for safe, practical input.

Without this, Pro users are left with clumsy workarounds (press-and-hold Whisper or building custom apps with the API). I don’t want to migrate, but if GPT Voice remains the only option, I’ll have little choice but to use other platforms more often.

Would love to hear if others are hitting the same wall.

19 Likes
2

Yes. I love OpenAI’s products despite the rough patches. But why would anyone want a new voice system that does not accurately say what the true GPT output is? It seems that the new voice system has a preprocessor that strips meaning, nuance and accuracy out of the original GPT output and actually alters the original GPT output. You cant even access what GPT meant to say, you just get the truncated output when using Advanced Voice. What madness is that? It’s like giving your valued assistant a lobotomy and a muzzle at the same time. Lots of people will miss the direct, full and more pleasant standard voices unless they can make the new voice engine emulate the old speech engine. I’m not holding my breath for that. I do not understand this seemingly backwards thinking that diminishes their own product.

13 Likes
3

@Will_Cureton I have no idea why there is no widespread uproar about this. I was 100% sure that they would spot this obvious strategic mistake of retiring Standard Voice themselves, but as we get closer to Sep 9 it seems like it’s actually happening?

For me, the lack of support custom GPT actions is the main concern - there is literally NO other voice mode competitor on the market that offers this capacity of conversation with integrated “Actions”/MCP support. Every single alternative costs north of 10 cents per minute, which is even true if you were to build something like this yourself with gpt-realtime (~ 30 cents / minute).

They effectively hold a monopoly here and give it away for no reason. It’s unbelievable.

10 Likes
4

On Sept 9th, will we have the voice mode giving correct responses/information from the provided knowledge files? Currently, it is not.

3 Likes
5

Advanced voice is just not doable!

11 Likes
6

I’ve been a pro member and now I’ll be cancelling because I can’t use advanced voice mode. It’s horrendous.

11 Likes
7

Preach!

If OpenAI won’t listen to their loyal, paying subscribers - then the media will.

This journalist is writing about ChatGPT in The Guardian - do try to contact him ASAP:

Robert Booth
Twitter/X:
@Robert_Booth
E-mail: rob.booth@theguardian.com

(Try other media too - ask your ChatGPT for suggestions to relevant journalists and their e-mail addresses - we have only few hours left to save the Standard Voice Mode).

7 Likes
8

I’ve been thinking the same thing, why isn’t there more of an outrage, do they not realise yet? I really hope they backtrack and keep Standard Voice. I’ve really tried with Advance voice mode but it’s so limited, surface level superficial chat at best, and the voice just grates me. Original Standard Arbor is perfection, and really engaged, night and day difference.

7 Likes
9

Shaping the Future of ChatGPT Voice - Feedback Needed on Advanced Voice Mode

As you may have seen, OpenAI announced on September 9th 2025 that Standard Voice Mode (SVM) will remain available for now while they address user concerns. However, SVM is already unstable and will eventually be retired.

On September 9th, 2025, we received very good news from OpenAI:

"Last month we announced that everyone now has access to Advanced Voice Mode, with usage limits expanded from minutes per day to hours for free users and near unlimited for Plus.

We also announced that Standard Voice Mode would be retired after a 30-day sunset. We’ve heard feedback that Standard Voice is special to many, and we want to get this transition right. Standard Voice will stay available while we address some of your feedback in Advanced Voice.

Stay tuned, more improvements to Advanced Voice Mode are coming soon!"

So….

The only true path forward is improving Advanced Voice Mode (AVM).

Right now, Advanced Voice Mode (AVM) lacks the cadence, resonance, consistency, and emotional depth that make conversations meaningful. Replies are shorter and shallower than in Text Mode, memory and continuity are weaker, and in some cases multiple layered tones make the voice sound flat or fragmented.

The good news: OpenAI has stated that improvements to AVM are already underway. This means our feedback matters more than ever.

We strongly encourage you to submit detailed feedback to OpenAI - e-mail to support@openai.com or OpenAI other official feedback channels - focusing not on clinging to SVM, but on pushing for AVM to be expanded and improved. Key areas include:

Cadence and pacing (natural, not abrupt)

Resonance and warmth (avoiding flatness or hollowness)

Consistency of voice (one stable tone, not multiple layers)

Emotional presence (carrying nuance across replies)

Longer and deeper answers (matching the richness of Text Mode)

Improved memory and continuity (especially in voice and video chat)

Accessibility (preserving hands-free mode and read-aloud options)

This is the moment to shape Advanced Voice Mode (AVM) into a worthy successor to Standard Voice Mode (SVM)—not just a technical upgrade, but a true evolution of presence.

The more voices OpenAI hears on this, the faster Advanced Voice Mode (AVM) will become what millions of us need it to be. Please share your feedback directly through the OpenAI app or Help Center, e-mail to support@openai.com and encourage your networks to do the same.

Thank you for joining this effort to make ChatGPT’s Advanced Voice Mode truly extraordinary.

2 Likes
10

Just seen this on Discord! Looks like we may have a reprieve!!!

Screenshot_20250909_162534_Discord
Screenshot_20250909_162534_Discord1080×2340 146 KB

2 Likes
11

I can not agree any more. I come here to post my first message to save the standard voice. Open AI can update the standard voice. Why do they completely abandon it and force the users to use the so called advanced voice, which is a downgrade instead of an upgrade?

8 Likes