- AI Fire
- Posts
- 🔥 $20 Coding War: Sam Altman’s "Drunk-Dunks" and Anthropic’s 2% Blunder
🔥 $20 Coding War: Sam Altman’s "Drunk-Dunks" and Anthropic’s 2% Blunder
Anthropic faces backlash after Claude pricing test. OpenAI responds with Codex staying on $0 and $20 plans as Sam Altman mocks the situation publicly.

Key Points
Anthropic triggered a developer riot after a pricing experiment appeared to move Claude Code from the $20 Pro plan to the $100 Max tier.
The company later walked back the changes, claiming the test only affected 2% of new sign-ups, but the 98% confusion forced a total reversion of documentation.
Sam Altman seized the PR opening, trolling Anthropic on X (formerly Twitter) with a blunt "ok boomer" and admitting he was sipping drinks during the rival’s meltdown.
OpenAI reaffirmed its commitment to Codex, promising to keep the tool on $0 and $20 plans to contrast with Anthropic's rising infrastructure costs.
Nothing smells like blood in the water quite like a pricing experiment gone wrong. On April 22, 2026, Anthropic accidentally triggered a developer riot by updating its docs to suggest Claude Code was moving from the $20 Pro plan to the $100 Max tier.
While Anthropic scrambled to call it a small test, Sam Altman took the opportunity to pour a few drinks and start a social media fire.
The Golden Child of AI safety just got a harsh lesson in community management, while OpenAI positioned itself as the people’s provider.
1. The 2% Blunder: Anthropic’s Pricing "Rug-Pull"
The chaos started when eagle-eyed developers noticed a quiet update to Anthropic’s pricing page. Suddenly, Claude Code, the tool that has redefined engineering in 2026, was no longer listed as a perk for the $20 Pro plan.
Anthropic’s Head of Growth, Amol Avasare, eventually clarified that the $100 floor was only being tested on 2% of new users. Because the public-facing documentation was updated for everyone, the existing user base assumed their $20 subscriptions were being nerfed.
By Tuesday evening, Anthropic had reverted all changes, but the damage to their safety and transparency brand was already done.
For those of us tracking the cost of compute, this wasn't just a typo. Anthropic admitted that when they launched their subscription tiers, they didn't foresee the Agentic Era.
Agentic Drain: Running an agent for 8 hours straight (like Claude Code) costs significantly more than a standard chat. Anthropic's margins are clearly under pressure.
The Pivot: Anthropic is now openly looking at different options to keep the lights on, signaling that the era of cheap, unlimited agents might be coming to a close for the smaller labs.
2. "OK Boomer": Sam Altman’s Drunk-Dunking
While Anthropic was in full damage-control mode, OpenAI’s leadership was having a party. Sam Altman didn't just point out the mistake; he turned it into a cultural moment that painted Anthropic as the out-of-touch incumbent.
In response to Anthropic’s explanation of the pricing test, Altman simply replied with "ok boomer." Shortly after, he posted: "tongiht i have had a couple of drinks," while leaning into the spectacle.
It was a rare, raw look at the CEO of a nearly trillion-dollar company enjoying a rival's self-inflicted wound.
OpenAI’s team flooded social media to promise that Codex will remain available on Free and Plus ($20) plans. They explicitly noted, "We have the compute and efficient models to support it," a direct shot at Anthropic’s apparent struggle to manage the costs of its own popularity.
Our Take: The Compute Moat vs. The Safety Brand
OpenAI knows that while Anthropic has been winning the Reasoning war, they can win the Economic war. By keeping Codex at $20, OpenAI is effectively predatory pricing Anthropic out of the market. They are betting that their massive Microsoft-backed infrastructure can lose money longer than Anthropic can stay altruistic.
While Anthropic attempted a surgical (if clumsy) test to find a sustainable price for agents, OpenAI chose to weaponize the confusion to win back the vibe.
If you are interested in other topics and how AI is transforming different aspects of our lives or even in making money using AI with more detailed, step-by-step guidance, you can find our other articles here:
Finally, We SOLVED the Missing Piece With ALL AI Videos: Custom Voice + Perfect Lip-Sync Guide*
You're Ambitious but Unfocused? AI Turns You Into a High-Leverage Generalist in 2026
Why I Switched From ChatGPT to Claude Without Losing a Single Thing (Full Migration Guide)
Beginner's Guide: Build Your First AI Voice Agent (With No Code)*
Stop Writing Proposals: Turn Sales Calls Into Decks in 3 Minutes With AI (Free Template)*
*indicates a premium content, if any
Reply