Liwati menyang isi utama
OpenAI

Riset

Produk
Build more natural voice experiences with GPT‑Live‑1 in the API

GPT‑Live‑1 brings natural, full-duplex voice conversations to the API, with stronger instruction following, custom voices, and telephony support.

Riset
An OpenAI model proposes a solution to the Navier–Stokes problem

We’re sharing an AI-generated solution to the Navier–Stokes Millennium Prize Problem, including a writeup and a formal proof in Lean.

Riset
Research acceleration: The view inside OpenAI

Inside OpenAI, coding agents are reshaping AI research. Explore early data on agent usage, experiment velocity, task complexity, and research acceleration.

Riset
GPT-6 Astra: Generasi anyar saka plihan kecerdasan

Ngenalaké GPT-6 Astra, model kita sing paling cerdas lan paling selaras nganti saiki, kanthi kemampuan paling mutakhir ing penggunaan komputer, ngoding, keamanan siber, lan sains.

Keselamatan
GPT‑6 Astra System Card

Astra is our first model to reach the Critical level of cybersecurity capability under our Preparedness Framework.

(mbukak ing jendhela anyar)
Keamanan
Insiden Hugging Face lan langkah sabanjuré

OpenAI nuduhake temuan saka insiden keamanan Hugging Face lan langkah-langkah sing ditindakake kanggo nguwatake keamanan, pemantauan, lan penyelarasan model AI.

Perusahaan
Laju pangembangan model ing jaman kapabilitas siber kritis

OpenAI nguwatake pangawasan, keselarasan, lan keamanan kanggo model AI tercanggih. Delengen carane pangreksan anyar ngarahake laju pangembangan model.

Publikasi
Sepuluh kemajuan ing matématika lan èlmu komputer téoritis

OpenAI nuduhake asil anyar kanggo masalah lawas ing matématika lan èlmu komputer téoritis, kalebu kemajuan ing géometri, kriptografi, lan kompleksitas.

Riset
Carane rong setelan ngundhakake skor tolok ukur ARC-AGI-3 nganti ping telu

Rong setelan API ningkatake kinerja GPT-5.6 ing ARC-AGI-3, kalebu skor lan efisiensi, kanthi nyimpen nalar lan nguripake pemadhetan.