Creator of ChatGPT, GPT-4, DALL-E, and Sora. Foundation model platform and developer APIs.
A Ramp customer story documenting Codex with GPT-5.5 for code review, with claimed reviewer-feedback turnaround dropping from hours to minutes. Reads as marketing collateral, but fits the pattern of role-by-role Codex case studies OpenAI has been publishing in batches.
View source ↗A research-side proof point: an OpenAI model is credited with solving the 80-year-old unit distance problem, disproving a long-standing conjecture. Lands inside the broader push to position GPT-5.5-era systems as frontier, not just productized.
View source ↗Next phase of the Education for Countries program — more partnerships, teacher training, classroom tools. Incremental scaling of an existing initiative rather than a new product line.
View source ↗Singapore joins the country-partnership roster with a multi-year deployment, talent, and public-services agreement. Continues the sovereign-AI deal cadence visible across Malta, Education for Countries, and prior rollouts.
View source ↗Bundles Content Credentials, SynthID, and a public verification tool into a stated provenance program. Moves OpenAI from talking about authenticity to shipping the surface that regulators and platforms have been asking for.
View source ↗Dell partnership brings Codex into hybrid and on-premise enterprise environments. Pulls the agent runtime out of OpenAI's cloud and into customer-controlled infrastructure — a structural shift in how the product is delivered.
View source ↗