
The fight between OpenAI and Anthropic for control of AI-assisted coding has turned into open commercial warfare. OpenAI is offering enterprises two months of free Codex usage to switch off rival coding tools — chiefly Anthropic's Claude Code — and it has paired the incentive with a one-click migration tool designed to make defecting almost frictionless. Anthropic responded within hours. The result is one of the most direct poaching campaigns the AI industry has seen, aimed squarely at the product that has become Anthropic's fastest-growing business.
For an engineering leader, the practical news is a rare window of leverage: two rivals are briefly competing to give your team more for less. The deeper news is that the lock-in that used to keep coding tools sticky may be evaporating.
What Is OpenAI Offering?
On May 13, 2026, OpenAI launched a 30-day "Switch to Codex" promotion: enterprises that apply within the window receive two months of free Codex enterprise usage. It targets business and enterprise customers migrating from competing platforms, and the credits cover net-new users — developers not already on a ChatGPT Business or Enterprise plan. Companies apply through a dedicated form on OpenAI's site, with the window closing around mid-June.
The "net-new users" condition is the catch to read carefully. The free usage rewards expansion rather than subsidizing a company's existing OpenAI seats, which makes the offer most valuable to organizations bringing a fresh cohort of developers onto Codex — precisely the developers that might otherwise have gone to Claude Code.
Why the Migration Tool Is the Real Weapon
An incentive alone rarely moves enterprises, because switching coding tools is high-friction: teams accumulate custom configurations, scripts, and workflows that are painful to rebuild. OpenAI's sharper move attacks that friction directly. Codex now includes a built-in migration tool: a user opens Settings, clicks "Import other agent setup," and Codex scans the machine for Claude Code's configuration — skills, hooks, MCP (Model Context Protocol) servers, subagents, instruction files, and up to 30 days of session history — then migrates automatically what it can. For anything it cannot convert cleanly, it opens a follow-up thread where an AI assistant walks the user through the rest.
Third-party utilities have formed around the same idea; an open-source tool called claude2codex converts a Claude Code setup in a single command, with developers reporting that most settings transfer and work on the first try. These tools generally classify each piece of a configuration by difficulty — fully mechanical conversions, items needing manual review, and pieces requiring a genuine rewrite — so teams can see what survives the move. The strategic point: vendor lock-in has been one of the main things protecting Claude Code's lead, and by making a team's accumulated investment portable in a few clicks, OpenAI is trying to dissolve the very switching cost that keeps customers from leaving.
Read more: Anthropic Ends Subscription Subsidy for Agents June 15: Credit Pool Replaces Flat-Rate Access
Why Is Anthropic the Target?
OpenAI is not attacking a peripheral product. Claude Code has been Anthropic's fastest-growing offering and a major driver of its enterprise revenue, and AI coding agents are one of the few categories where AI delivers clear, measurable productivity gains. Winning developers is also strategically compounding: the platform a company's engineers code on shapes which models it buys, which APIs it standardizes on, and how deeply it integrates one vendor. Capturing a developer team is a foothold into an enterprise's entire AI stack.
That is why OpenAI will give away two months of a premium product. In a market where switching costs and developer habits create durable lock-in, the economics favor buying share now and monetizing later — the same land-grab logic that drove earlier platform wars.
How Did Anthropic Respond?
Anthropic did not wait. Within hours of the Codex promotion, it raised Claude Code weekly usage limits by 50% through July 13 — a direct move to blunt the appeal of switching by giving existing customers more for the same price. It is also reworking how it charges for agents, ending its flat-rate subscription subsidy and shifting to a credit-pool model, a change that reflects the brutal compute economics beneath these tools, where heavy agentic use can cost a provider far more than a flat subscription brings in.
That billing reality is the uncomfortable backdrop to the whole fight. Coding agents consume enormous compute, and both companies subsidize usage to win share while hunting for a price structure that does not lose money on power users. OpenAI's free-Codex offer and Anthropic's quota expansion are, at bottom, two bets on how to spend that subsidy: OpenAI on acquisition, Anthropic on retention.
Read more: Salesforce Will Not Hire More Software Engineers Next Year As Claude Code Compresses Migrations
What Should Enterprise Buyers Do?
For engineering leaders, the price war is leverage. Two months of free Codex plus near-automated migration lowers the cost and effort of at least trialing a switch, while Anthropic's enlarged limits make staying more attractive — both vendors are, for now, competing to give customers more. The caveats matter: the free Codex usage applies to net-new users, not existing seats; migration tools move most but not every configuration, and anything classified as a rewrite needs hands-on work; and choosing a coding platform is a multi-year commitment that outlasts a two-month promotion. A tool that is free for two months but a worse fit for a team's workflow is not a bargain.
The deeper takeaway is that the moat around AI coding tools is thinner than it looked. If a team's entire Claude Code setup can be ported to Codex in a few clicks, neither company can rely on lock-in to hold customers — they have to keep winning on the model, the experience, and the price. That is good for buyers and punishing for margins, and it explains why this front in the AI wars turned so aggressive so fast. This article is not investment advice.
Frequently Asked Questions
What is OpenAI's Codex migration offer?
OpenAI launched a 30-day "Switch to Codex" promotion on May 13, 2026, giving enterprises that apply two months of free Codex usage to move from rival coding tools like Claude Code. The free usage covers net-new users — developers not already on a ChatGPT Business or Enterprise plan — and applications run through a form on OpenAI's site.
How does the Claude Code to Codex migration tool work?
In Codex Settings, users click "Import other agent setup," and Codex scans for Claude Code's configuration — skills, hooks, MCP servers, subagents, instruction files, and up to 30 days of session history — migrating what it can automatically. An AI-assisted thread handles the rest, and open-source tools like claude2codex offer a one-command alternative.
How did Anthropic respond?
Within hours, Anthropic raised Claude Code weekly usage limits by 50% through July 13 to discourage switching, while also moving agent billing from a flat-rate subscription to a credit-pool model. The dueling moves reflect a price war driven by the high compute cost of running coding agents.
Should an enterprise switch from Claude Code to Codex?
It depends on fit, not just the promotion. The free offer and easy migration make trialing Codex low-cost, but the credits apply only to net-new users, not every configuration transfers cleanly, and a coding platform is a multi-year decision. Teams should weigh model quality and workflow fit over a two-month discount. This article is not investment advice.
ⓒ 2026 TECHTIMES.com All rights reserved. Do not reproduce without permission.




