Anthropic is seemingly making rapid progress with Claude. The company recently announced Claude Mythos, an AI so good that it can hack just about any major web browser, or software. At the same time, Anthropic’s rival, OpenAI, was celebrating a milestone – reaching 3 million users for its coding agent Codex. OpenAI CEO Sam Altman announced that the company was resetting all usage limits for Codex for achieving this feat.
On X, Sam Altman wrote, “To celebrate 3 million weekly codex users, we are resetting usage limits.” The OpenAI chief stated that the company will reset the limits for every 1 million new users, until it reaches 10 million users for Codex.
Do note that the limit reset essentially gives you clean sheet access for using the AI tool for the remainder of the period. That is, if you had used 20 per cent of your monthly limit, now you can once again start from 0.
Claude Code chief calls Mythos terrifying
Claude Code creator Boris Cherny wrote on X that Claude Mythos was so good that it was almost scary. He wrote, “Mythos is very powerful, and should feel terrifying.” Cherny insisted that the company was taking steps to work on cybersecurity.
Do note that Mythos was intended to originally be a coding tool, likely the next version of Claude Code.
Anthropic stated that Mythos was able to find thousands of vulnerabilities across various softwares, that were missed by humans. The company is not planning to give access to Claude Mythos to the public, and was instead working with companies like Google and Microsoft to fix security vulnerabilities Mythos may find, as part of Project Glasswing.
Additionally, Anthropic recently stated that its run-rate revenue (RRR) has now crossed $30 billion, up from $9 billion at the end of 2025. On the other hand, OpenAI, which is expected to go public before the end of this year, is said to have an RRR of $25 billion as of March 2026.
This could raise concerns within the Sam Altman-led firm, which is currently valued at $852 billion, as compared to $380 billion valuation for Anthropic.
OpenAI getting more Codex users
Codex was Anthropic’s answer to Claude Code. The Sam Altman-led firm has doubled down on its coding agent as it shifts focus on enterprise products. According to Codex head Thibault Sottiaux, OpenAI managed to get 1 million Codex users in less than a month.
He wrote on X, “Three million people are now using Codex weekly – up from two million a little under a month ago. Incredible to see the growth.”
Tibo had previously reset limits for Codex on April 1, 2026 after he found that “increased rate of users hitting rate limits.”
Around the same time, many Claude users too claimed that they were hitting limits faster than before. However, after investigating the issue, Anthropic later said that no user was charged unfairly.
