OpenAI Codex Hits Five Million Weekly Users as Non-Coders Drive Growth

AI coding assistant Codex has hit a striking milestone, with OpenAI announcing more than five million weekly active users. The tool, which started as a developer-focused coding agent built on a specialized version of the o3 model, is now expanding rapidly into broader knowledge work including research, data analysis, and report generation.

From Code to Knowledge Work

When OpenAI launched the Codex desktop app in February 2026, it served primarily as a tool for software developers. Just four months later, the weekly user base has grown more than six times over. Perhaps more striking is who is driving that growth. Knowledge workers — professionals in fields like finance, marketing, and operations — now make up roughly 20 percent of Codex users and are adopting the tool more than three times faster than developers.

These non-technical users are putting Codex to work in ways that go well beyond writing code. Around 72 percent of knowledge worker users produce artifacts each week, including reports, memos, contracts, spreadsheets, and PDFs. The fastest-growing use cases paint a clear picture: data analysis surged 110 percent week over week, while research tasks grew 37 percent and knowledge artifact creation climbed 36 percent.

Parallel Processing Changes How People Work

One of the most notable shifts is how people interact with Codex. About 50 percent of users now run multiple tasks simultaneously, up from less than a third in mid-April. This parallel capability — where a large language model, or LLM, handles several requests at once — lets professionals orchestrate complex workflows without waiting for each step to finish before starting the next.

Enterprise adoption has also accelerated significantly. Companies like Cisco, Nvidia, and Rakuten have deployed Codex across their organizations, with Cisco reporting a 50 percent reduction in code review times. Business and enterprise tier usage has grown six times over since January 2026, and the tool’s npm downloads jumped from 82,000 in April 2025 to 14 million by March 2026.

The rapid expansion signals a broader trend in AI productivity tools: these systems are no longer just for engineers. As knowledge workers discover they can automate research, generate reports, and analyze data at scale, the line between developer tools and everyday business software continues to blur.