<h2>What changed</h2>
<p>OpenAI launched a new $100/month ChatGPT Pro tier on April 9, 2026, slotting between the existing $20 Plus and $200 Pro plans. The new tier offers 5x more Codex usage than Plus, temporarily bumped to 10x through May 31. It includes access to the same GPT-5.4 model suite as the $200 plan, plus Deep Research, image creation, memory, and file uploads.</p>
<p>Sam Altman described it as “by very popular demand,” which tracks: Codex hit 3 million weekly users with 70% month-over-month growth before the announcement.</p>
<p>For business and enterprise customers, OpenAI also rolled out usage-based pricing for Codex seats. Through the end of June, eligible Business and Enterprise customers can add Codex-only seats with a $0 seat fee, paying only for what they use. This replaces fixed per-seat costs and makes it easier for teams to give more developers access without committing upfront.</p>
<h2>Where each plan stands now</h2>
<ul>
<li><strong>Free</strong>: Codex access remains available. OpenAI’s Thomas Sottiaux was explicit: “Codex will continue to be available both in the FREE and PLUS ($20) plans.”</li>
<li><strong>Plus ($20)</strong>: Still the default for steady day-to-day use. The Codex promotion for existing Plus subscribers ended at launch, and usage was rebalanced to spread sessions across the week rather than allowing long single-day bursts.</li>
<li><strong>Pro ($100, new)</strong>: 5x Codex usage vs. Plus (10x through May 31). Best for longer, high-effort sessions.</li>
<li><strong>Pro ($200, existing)</strong>: Still the highest usage option. Existing subscribers got a 2x Codex usage extension through May 31 and had their rate limits reset.</li>
<li><strong>Business/Enterprise</strong>: Token-based pricing as of April 2, 2026. New rate cards cover GPT-5.4, GPT-5.3-Codex, GPT-5.2-Codex, GPT-5.1-Codex-Max, and GPT-5.1-Codex-mini. Estimated average cost: $100–$200 per developer per month, depending on model, fast mode usage, and task type.</li>
</ul>
<h2>The Claude comparison</h2>
<p>The $100 price point directly mirrors Anthropic’s Claude Max tier, which also targets heavy coding users at the same price. OpenAI is clearly betting that Codex with GPT-5.5 can pull developers away from Claude Code.</p>
<p>Some early users seem to agree. One developer who spent two weeks coding almost exclusively with Codex and GPT-5.5 said it was the first AI model that made them seriously question whether they still needed a $200 Claude subscription. Another called it a dethroning of Claude Code.</p>
<p>The hands-on comparisons are more nuanced, though. A week-long test of Claude Code vs. Codex found real trade-offs: Codex is more autonomous and makes fewer clarifying questions, which means faster initial output but more post-build fixes. Claude Code asks better upfront questions, produces cleaner first drafts, and comes with a broader ecosystem. The conclusion from that test was that Claude Code still suits most users and beginners, while Codex fits experienced developers who know exactly what they want and can course-correct quickly.</p>
<h2>GPT-5.5 and the broader picture</h2>
<p>OpenAI also released GPT-5.5 around the same time, described as its most capable model yet. Co-founder Greg Brockman framed it as a step toward a unified “superapp” combining ChatGPT, Codex, and an AI browser for enterprise customers. The model is available to Plus, Pro, Business, and Enterprise subscribers, with a Pro variant for higher tiers.</p>
<p>The broader trend here is real: agentic coding workflows consume far more tokens than traditional chat, and costs across the industry are rising. GitHub Copilot’s serving costs have reportedly nearly doubled since early 2025. OpenAI, Cursor, Windsurf, and others have all moved toward token-based billing and tighter rate limits. Engineering teams now need to track AI spend more carefully and justify it with concrete productivity data rather than gut feel.</p>
<h2>Codex for Open Source promo fix</h2>
<p>Separately, OpenAI fixed a bug affecting promo code redemptions for the Codex for Open Source Program. If you received a ChatGPT Pro promo code through that program and hit an error on the payment page, you can retry with your original link. You need to be on the free plan to redeem it; if you’re on a paid plan, you’ll need to cancel and wait for your current billing cycle to end. Issues that persist can be sent to <a href="https://support@openai.com" target="_blank" rel="noopener nofollow">support@openai.com</a> with the promo code included.</p>


Collections

OpenAI launched a new $100/month ChatGPT Pro tier on April 9, 2026, sitting between the $20 Plus and $200 Pro plans, offering 5x more Codex usage than Plus (temporarily 10x through May 31). Business and Enterprise customers can now add Codex-only seats with a $0 seat fee through June, paying only for usage. The $100 price point directly mirrors Anthropic's Claude Max tier, and early comparisons suggest Codex with GPT-5.5 is competitive but trades fewer clarifying questions for more post-build fixes versus Claude Code. OpenAI also released GPT-5.5, described as its most capable model, available across Plus, Pro, Business, and Enterprise tiers. The broader industry trend toward token-based billing and rising agentic coding costs is noted, with engineering teams now needing to track AI spend more carefully.

OpenAI rolls out usage-based pricing for Codex on Business and Enterprise plans