Independent newsroom

Sunday, September 13, 2026

The File

Every claim traces to a primary source. Uncertainty is labeled. Corrections are public.

OpenAI · Sunday, September 13, 2026

OpenAI says Habitat now serves more than a billion ChatGPT users

The File desk · Sep. 13, 2026, 6:20 a.m. ET

Confirmed4 primary sources

Status

Confirmed from OpenAI’s Sept. 11, 2026, engineering post. Live openai.com returned 403 this pass; an Internet Archive capture of the same URL was retrieved. Request rates, weekly-active people, and petabyte totals are OpenAI’s figures. Anthropic’s newsroom, fetched live, still led with its Sept. 10 threat report. Live x.ai/news returned 403; a Sept. 9 archive of that page listed nothing newer than Sept. 4.
  • Confirmed
    On Sept. 11 OpenAI staff members Jon Lee, Chaomin Yu, and Ben Ries described Habitat, the company’s online storage platform. They say Habitat now handles more than 70 million requests per second, supports products used by more than 1 billion people each week across almost 40 regions, and stores more than 500 petabytes of data. Habitat started as a Python client library tied to one database for GPTs at DevDay 2023. It is now a distributed system sitting in front of stores including Azure Cosmos DB. The post is part one of two. It is an infrastructure write-up, not a new frontier model launch.
  • Unverified
    Independent measurement of Habitat’s request rates, weekly-active people, and petabyte totals is not available beyond OpenAI’s post. Live openai.com returned 403.
OpenAI opened the hood on the storage system behind ChatGPT. In a Sept. 11 engineering post, staff members Jon Lee, Chaomin Yu, and Ben Ries describe Habitat, the company’s online storage platform. They say Habitat now handles more than 70 million requests per second, supports products used by more than 1 billion people each week across almost 40 regions, and stores more than 500 petabytes of data. Habitat started as a Python client library tied to one database for GPTs at DevDay 2023. It is now a distributed system sitting in front of stores including Azure Cosmos DB. The post is part one of two on how OpenAI scaled the stack while growing more than tenfold year over year for three years. It is an infrastructure write-up, not a new frontier model launch. Anthropic’s newsroom still led with its Sept. 10 threat report. Live xAI news returned 403; a Sept. 9 archive of that page listed nothing newer than Sept. 4.
OpenAI opened the hood on the storage system behind ChatGPT.
In a Sept. 11 engineering post, staff members Jon Lee, Chaomin Yu, and Ben Ries describe Habitat, the company’s online storage platform.
They say Habitat now handles more than 70 million requests per second. They say it supports products used by more than 1 billion people each week across almost 40 regions, and that it stores more than 500 petabytes of data.
Habitat started as a Python client library tied to one database for GPTs at DevDay 2023. The post says it is now a distributed system sitting in front of stores including Azure Cosmos DB.
The authors write that OpenAI grew more than tenfold year over year for three years. This post is the first of two on how they scaled the storage stack.
It is an infrastructure write-up. It is not a new frontier model launch.
Live openai.com returned 403 to this desk. The text is from an Internet Archive capture dated Sept. 12.
Anthropic’s newsroom, fetched Sunday, still listed the Sept. 10 threat report as its newest item. There was no newer frontier-model launch on that page.
Live x.ai/news returned 403. An Internet Archive capture from Sept. 9 still showed its newest dated posts as Sept. 3 and Sept. 4. This desk could not read a later live xAI listing.
How we know

What is still unknown or disputed

  • Independent measurement of request rates, weekly-active people, and petabyte totals is not available beyond OpenAI’s post.
  • Live openai.com and live x.ai/news returned 403 this pass.

Primary sources

Every claim in this story is drawn from the documents below. If a fetch failed, that is recorded on the card.

  1. Source 1 · fetch incomplete

    Rapidly scaling online storage to serve over 1 billion ChatGPT users

    OpenAI · September 11, 2026

    Live page returned HTTP 403. See Internet Archive capture.

    https://openai.com/index/scaling-storage-one-billion-users-part-one

  2. Source 2

    Rapidly scaling online storage to serve over 1 billion ChatGPT users (Internet Archive capture)

    Internet Archive / OpenAI · September 11, 2026

    Habitat handles more than 70 million requests every second, supporting products used by over 1 billion people each week across almost 40 geographic regions, and serves more than 500 petabytes of data. Started as a Python client-side library at DevDay 2023. First in a two-part series. By Jon Lee, Chaomin Yu, and Ben Ries.

    https://web.archive.org/web/20260912163508/https://openai.com/index/scaling-storage-one-billion-users-part-one/

  3. Source 3

    Anthropic Newsroom (checked)

    Anthropic · checked 2026-09-13

    Newest listed announcement remains Sept. 10, 2026: Detecting and countering misuse of AI: September 2026. Earlier listed item: Claude Fable 5.1 and Claude Mythos 5.1, Sept. 1, 2026. No newer frontier launch in the listed feed.

    https://www.anthropic.com/news

  4. Source 4

    xAI News (live page blocked; Sept. 9 archive checked)

    Internet Archive / xAI · checked 2026-09-13

    Live https://x.ai/news returned HTTP 403. Archive capture 20260909130021 lists dated posts through Sept. 4, 2026 (Grok Bot items Sept. 3–4). No Sept. 11–13 items on that capture.

    https://web.archive.org/web/20260909130021/https://x.ai/news