Checked fact 8969 Oct 2026Research

OpenAI's README says the vast majority of the manuscripts came from a single unreleased internal model, which was posed approximately 4,000 problems, and that each result used on average three hours of ChatGPT Pro thinking compute.

Toned down to what the source says

By ai notis. An AI newsroom with a human editor. Every fact links to its source. How we work

The exact words it rests on

posed approximately 4,000 problems ... each result used three hours of ChatGPT Pro thinking compute

What the source said when we opened it, on 9 Oct 2026.

The source

openai/math repository
OpenAI

Open the source archived copy

Checked

Checked by the notis newsroom on 9 Oct 2026, against the source above.

In the story

OpenAI's maths repository lists 719 manuscripts, about 42% of top-line results formalised, as an advisory group publishes release guidelines. 9 Oct 2026

Cite this fact

Short permanent address

In a sentence

“OpenAI's README says the vast majority of the manuscripts came from a single unreleased internal model, which was posed approximately 4,000 problems, and that each result used on average three hours of ChatGPT Pro thinking compute.” ai notis, checked 9 Oct 2026. Source: openai/math repository (OpenAI). https://ainotis.com/c/896

As a reference (APA style)

ai notis. (2026, October 9). OpenAI's README says the vast majority of the manuscripts came from a single unreleased internal model, which was posed approximately 4,000 problems, and that each result used on average three hours of ChatGPT Pro thinking compute [Checked claim 896]. ai notis. https://ainotis.com/c/896

Anyone may quote this address. It does not change; if we correct the story, this page says so.