Atlas record: https://www.aitimeline.dev/en/events/evt-0299/
Availability · May 13, 2024
GPT-4o
OpenAI released a model trained on text, audio and image at once; it answered in voice with a latency of about a third of a second.
Why it matters
Spoken conversation with a machine first ran at human tempo, because the three separate models that lost tone and timing between them were gone.
Until then voice mode was a recogniser, a language model and a synthesiser. Each stage added delay, and more importantly the recogniser passed on words alone: tone, pauses, laughter, several voices in the room never reached the model. A single network trained on audio directly hears all of it and can answer with intonation. The o in the name stands for omni. The model also became free for most ChatGPT users, putting the frontier within reach without payment for the first time.
Event record
- Event date
- May 13, 2024
- Timeline date
- Event date
- Verification
- Sources gathered automatically · September 18, 2026
- Lines
- ID
- evt-0299
Announcement of 13 May 2024.
Related events
Earlier
- Extends GPT-4 2023
The same line of models, now trained on text, audio and image together.
- Related Whisper 2022
A pipeline of three models was replaced by one, and conversational latency fell to human levels.
- Related WaveNet 2016
Generating sound directly, without an intermediate text stage.
Later
- Related Moshi 2024links to this record
The same year and the same idea of a single model for audio, but with open weights.
- Related GPT-5 2025links to this record
The previous attempt to gather everything into one model was about data formats; this one is about how to think.
- Extends gpt-realtime 2025links to this record
The same line of models working with audio directly, carried to production use.