Rapidly scaling online storage to serve over 1 billion ChatGPT users

OpenAI details how its Habitat storage system grew from a Python library into a globally distributed platform supporting 1 billion ChatGPT users and 22 million requests per second.

By AIWI EditorialReviewed by Chayanika DivekarAI-assisted summary · human-reviewed1 min readPublished Sep 11, 2026

Single source1 source

Reality Check

Availability: Shipping now
Hype-o-meter: Some hype
Who it's for: Infrastructure engineers · Enterprises
Do this now: Read OpenAI's write-up on Habitat's architecture if you build or operate large-scale storage systems.
Claims check: Scale figures (1 billion users, 22M requests/second) are stated by OpenAI in the input and not independently verified here.
Corroboration: Single source · 1 source

Key Takeaways

Habitat evolved from a Python library into a distributed storage platform
The platform serves over 1 billion ChatGPT users
It handles 22 million requests per second
The system is globally distributed

OpenAI has published an account of how it scaled its online storage infrastructure to support ChatGPT at large scale. According to the company, its storage system, called Habitat, began as a Python library and was rebuilt into a globally distributed storage platform. OpenAI states that the platform now serves over 1 billion ChatGPT users and handles 22 million requests per second.

Why it matters

Serving a user base of this size at high request volumes requires storage infrastructure that can operate reliably across global regions. OpenAI’s description highlights the engineering path from a single library to a distributed platform designed for this level of demand.

Who should care

Infrastructure and platform engineers, as well as teams operating large-scale online services, may find the account relevant to understanding how storage systems are adapted to meet high-throughput, high-user-count requirements.

Was this helpful?