OpenAI says its products now reach over 1 billion weekly users and 2.5 million businesses

OpenAI confirms the scale

OpenAI has confirmed that its products are now used by more than 1 billion people every week. According to a dev.to report on the company's September 2026 announcement, the figure covers OpenAI's entire product ecosystem — ChatGPT, Codex and related offerings — and the same post puts business adoption at 2.5 million organisations.

The confirmation follows months of speculation. Independent reporting in July and August 2026 suggested OpenAI was close to, or already past, the billion-user threshold, and the company has now put its own name to the number.

An ecosystem figure, not a ChatGPT figure

The distinction matters. OpenAI did not claim that ChatGPT alone passed 1 billion weekly active users; the milestone aggregates its consumer and developer products. ChatGPT is the centre of that picture, but Codex and other tools count toward the same total.

According to dev.to, OpenAI's announcement also points to continued compute and cost advantages that the company says will support further growth in usage. Notably, the post does not introduce new pricing, cheaper business tiers or specific integration features — so the economics signal is strategic rather than a concrete commercial offer.

A comparison to avoid

The dev.to report cautions against pairing OpenAI's number with a figure sometimes cited for Perplexity: 780 million queries in a single month. TechCrunch reported that number for May 2025, making it historical context rather than fresh news, and it counts queries rather than weekly active users. The two metrics measure different things and are not directly comparable.

What businesses should take from it

The dev.to article argues the headline number is less important than what it implies: AI assistants and coding tools are becoming ordinary workplace infrastructure. As exposure grows across employees, customers and suppliers, the organisational friction of adopting AI tends to fall.

The practical guidance from the report includes:

• Start with small, repetitive workflows — drafting, summarising, research, customer replies and code — rather than a company-wide rollout.
• Check whether AI tools can sit inside the systems staff already use, instead of forcing manual copying between applications.
• Watch actual pricing, usage limits and plan terms before committing budget; OpenAI's cost comments are a signal, not a published price change.
• Set clear review points for customer-facing, financial, legal or operational outputs where AI errors could cause harm.
• Treat Codex's inclusion in the metric as a reminder that AI-assisted software development is a central part of adoption, not a side effect.

The report also warns against drawing procurement conclusions from a reach metric. A marketing team, a support function and a software team may each need different tools and controls, and testing a specific workflow with clear ownership says more than any headline user count.

Why it matters

A billion weekly users across 2.5 million businesses means AI tools have moved well past the experimental phase for a large share of the working population. The competitive question is shifting from whether teams can access capable models to how well they implement them: retrieving the right information, triggering the right process and keeping humans in the loop where errors carry real cost. OpenAI's scale guarantees attention; it does not guarantee that any particular tool fits any particular job.