News · May 13, 2024
OpenAI launches GPT-4o: faster, half the API price, built for voice
OpenAI has introduced GPT-4o, a model that handles text, images and audio and costs half as much as GPT-4 Turbo in the API. For product teams, cheaper frontier-level calls change the math on many AI features.
What happened
- OpenAI announced GPT-4o on May 13, 2024 and said that in the API it is twice as fast as GPT-4 Turbo, half the price, and has rate limits five times higher, according to its announcement.
- According to OpenAI, the model can respond to audio in as little as 232 milliseconds, around 320 milliseconds on average; at launch, text and image input with text output were released publicly.
- OpenAI said GPT-4o improves markedly on non-English text and uses fewer tokens in many languages, citing 20 languages in its comparison.
- PCWorld reported that GPT-4o is being made available to free ChatGPT users with a usage limit, after which conversations fall back to GPT-3.5.
Why it matters
Halving the price of a top model at the same time as doubling speed makes some features viable that were too slow or too expensive a month ago, especially for apps with many users and short interactions. Better token efficiency outside English also matters for products sold in several markets.
We would rerun cost and latency estimates for existing AI features on the new model, test output quality on real user cases, and only then switch. Keep evaluations in place, because a cheaper model is only a saving if quality holds for your users.
Sources
This post is CMM’s own summary and commentary on publicly reported news. Facts are attributed to the sources linked above; please read them for full details. CMM is not affiliated with these publishers or with the companies mentioned, and trademarks belong to their owners. Prepared with AI assistance. Corrections or requests: [email protected]. See our editorial policy.
Want help with this?
Tell us what you need, and we will suggest the right team model on a short call.
Talk to us