ChatGPT May Get 8 New Voices. Here's How They Sound
OpenAI has yet to confirm or deny the existence of these voices.
By Sukhbir Cheema — 05 Sep 2024, 03:38 PM
OpenAI may soon add eight new voices to ChatGPT's Advanced Voice Mode
TestingCatalog, a tech news platform specialising in AI reporting, recently conducted reverse engineering and reported that OpenAI could soon add eight new voices to ChatGPT's current voice options.
Meanwhile, engineer Tibor Blaho has shared samples of each voice on X (formerly Twitter).
OpenAI has yet to confirm or deny the existence of these voices.
The alleged new voice options are Fathom, Glimmer, Harp, Maple, Orbit, Rainbow, Reef, Ridge, and Vale
These voices may be added to ChatGPT's current lineup, which includes Ember, Juniper, Cove, and Breeze. Notably, Sky, a previous voice option, was removed following Scarlett Johansson's allegation that OpenAI used her voice without permission.
The new voices reportedly feature a variety of genders and accents. Fathom and Vale have an upper-crust British tone, while Reef has a slightly Australian accent.
ChatGPT's Voice Mode was first announced in May as part of the GPT-4o launch, which introduced the new AI model that blends text, visuals, and audio modalities into one system.
In the demo released on X, the new ChatGPT voices were heard mimicking animal sounds, such as barking like a dog and cawing like a crow
While some voices replicated the animal sounds convincingly, others simply said the words "woof" and "caw".
That said, what's most impressive is how natural these voices sound.
ChatGPT's Voice Mode is currently only available to selected ChatGPT Plus users.
