Summary
OpenAI announced GPT-4o on May 13, 2024 and demonstrated its low-latency voice and multimodal capabilities. Text and image capabilities began rolling out in ChatGPT, including the free tier, and developers received text-and-vision API access. The new Voice Mode was scheduled for a later alpha for ChatGPT Plus, while new audio/video API capabilities were planned for a limited group of trusted partners.
Editorial correction, October 26, 2026
The primary announcement separates demonstrated capabilities from launch-day access. It places the new Voice Mode alpha and limited-partner audio/video API rollout in the coming weeks. The original broad availability language must not be read as general availability of the demonstrated new voice mode on May 13.
Revision record with the prior summary and metadata.
The original narrative below is preserved as published; use the revised summary and this note for the corrected account.
What Happened
On May 13, 2024, OpenAI launched GPT-4o at a Spring Update event. The "o" stood for "omni," reflecting the model's native ability to process and reason across text, audio, and vision simultaneously — rather than piping different modalities through separate models.
The most dramatic improvement was in voice interaction: GPT-4o could respond to audio input in as little as 232 milliseconds, approaching human conversational speed. The model could detect and respond to emotional tone, handle interruptions, and produce expressive speech output including laughter, singing, and dramatic delivery.
OpenAI made GPT-4o available to all ChatGPT users, including those on the free tier — a significant departure from the previous approach of reserving the most capable models for paying subscribers. The API was priced at half the cost of GPT-4 Turbo, making it both the most capable and the cheapest flagship model OpenAI had offered.
The live demo showcased the model tutoring a student in math via camera, translating conversations in real-time, and narrating scenes from a phone's camera feed. The voice capabilities were demonstrated by OpenAI researchers having a natural, flowing conversation with the model.
Why It Matters
GPT-4o represented a meaningful shift in what AI interaction could feel like. Previous AI assistants had perceptible latency, mechanical voices, and limited ability to handle the nuance of real conversation. GPT-4o's near-instant responses and expressive voice brought AI interaction closer to the science fiction ideal — and closer to replacing human interactions in customer service, tutoring, and companionship applications.
The decision to make GPT-4o free for all users was a competitive move that changed the market dynamic: previously, OpenAI's best models were behind a paywall, giving free-tier alternatives (like Google's Gemini in Search) a chance to capture users. By making GPT-4o free, OpenAI aimed to maintain its user base against growing competition.
The voice capabilities would generate controversy just days later, when comparisons to Scarlett Johansson's character in the film "Her" triggered a legal dispute over one of the demo voices.
§ How to read the metadata
- Landmark
- Fundamentally alters the trajectory; 2–5 per year.
- Major
- Meaningfully shifts the landscape; 2–4 per month.
- Notable
- Worth documenting; significance can be upgraded later.
- Confidence
- High = primary sources corroborate. Medium = credible secondary only. Low = provisional. Disputed = credible sources disagree.
- Contestation
- Uncontested = no formal challenge. Contested = at least one challenge open. Superseded = replaced by a later entry. Unresolved = dispute still open.