Overview

GPT-4o, announced in May 2024, was OpenAI’s first model trained end to end across text, vision and audio. Its “o” stands for “omni”. It brought GPT-4-level intelligence to free ChatGPT users and enabled low-latency voice conversations in which the model hears tone and can respond expressively.

Highlights

  • Native text, image and audio understanding
  • Low-latency voice interaction
  • 128K-token context window
  • Faster and cheaper than GPT-4 Turbo at launch
  • Later versions added native image generation in ChatGPT

Typical uses

Voice assistants, real-time conversation, image understanding, customer service and general-purpose chat applications.

Things to consider

  • Superseded by newer OpenAI models for many tasks.
  • Audio features in the API have separate pricing and endpoints.
  • Check model version identifiers, as GPT-4o received several updates.

Specifications are taken from the provider’s official announcement, documentation or model card. Benchmark results depend on evaluation settings; see how AI benchmarks work.