OpenAI Picks a Fight at $100/Month

For years, the gap between $20 and $200 was where OpenAI left developers to figure it out themselves. Not anymore.

What Just Happened

OpenAI announced a new $100/month plan on Thursday, slotting it squarely between its $20 Plus tier and the $200 Pro tier. The pricing ladder now looks like this: free (with ads), $8/month Go (with ads), $20/month Plus (no ads), $100/month — new — and $200/month Pro (still available, just quietly absent from the main pricing page).

The new tier is built around Codex, OpenAI's AI coding agent. The $100 plan delivers 5× more Codex usage than Plus. The $200 plan offers 20×. Neither is unlimited — a detail worth keeping in mind, especially since OpenAI is offering elevated limits on the $100 plan only through May 31. Developers who sign up, burn through code all day without hitting a wall, and assume that's the new normal should expect a reality check come June.

The usage numbers behind this launch are striking. OpenAI says Codex now has more than 3 million weekly active users globally, up 5× in the past three months, with month-over-month growth exceeding 70%.

The Real Story: This Is About Anthropic

Sponsored

Advertise with Us

[email protected]

Sponsored

OpenAI didn't bury the competitive angle. A spokesperson told TechCrunch directly: "Compared with Claude Code, Codex delivers more coding capacity per dollar across paid tiers." Naming a competitor in an official statement is a deliberate choice. This isn't a routine pricing update — it's a positioning move.

Anthropic has offered a $100/month option for Claude for some time, and Claude Code has built a loyal following among developers who value its reasoning depth and context handling. The timing of OpenAI's launch is pointed: Anthropic recently announced that Claude Code subscribers would need to pay extra for certain usage going forward — effectively a price increase for existing users. OpenAI moved fast to offer an alternative at the exact same price point.

The message is blunt: if you're reconsidering your Claude Code subscription, here's a door.

Three Ways to Read This

For individual developers, more options at a meaningful price point is genuinely useful. The $100 tier fills a real gap — Plus limits have frustrated power users for months, and $200 felt steep for anyone not running an enterprise workflow. But the temporary elevated limits are a classic land-and-lower move. Developers who stress-test the plan before May 31 should do so with that caveat front of mind.

For enterprise buyers, the calculus is shifting. AI coding tools are no longer a nice-to-have — they're becoming a line item. At $100/developer/month, a team of 20 runs $24,000/year, before any enterprise negotiation. That's real budget, and it puts Codex in direct competition with GitHub Copilot Enterprise, Cursor, and others. Procurement teams will want to see productivity benchmarks, not just marketing claims about "coding capacity per dollar."

For investors watching the AI tools space, this price war signals something structural. When two leading AI labs compete on cost-per-output rather than raw capability, it suggests the underlying models are converging. Differentiation is moving toward UX, integrations, and reliability — not just model quality. That's a different kind of moat.