OpenAI adds $100 Pro tier
OpenAI introduced a new $100 Pro plan aimed at heavier coding workflows and quietly upgraded the fallback model users see when rate limits hit, changing how intensive users access Codex. The plan sits between the $20 Plus tier and the $200 Pro tier and reportedly offers about five times the Codex usage of Plus, seeking to segment users by workflow intensity rather than broad categories. That pricing shift and a smarter backup model change the subscription ladder and create clearer upgrade and churn questions for product teams and revenue models. (economictimes.indiatimes.com; techradar.com)
OpenAI used to have a strange gap in ChatGPT pricing: $20 for Plus, then a jump all the way to $200 for Pro. On April 9, OpenAI inserted a new $100 Pro tier between them and said it is built for people doing longer Codex coding sessions. (community.openai.com) Codex is OpenAI’s coding agent, which means it is the part of ChatGPT built to help write, edit, review, and ship software. OpenAI describes it as a tool for features, refactors, migrations, reviews, and releases, not just quick code snippets. (openai.com) The new $100 plan is not replacing the old $200 plan. OpenAI’s announcement says the $100 tier gets about 5 times the Codex usage of Plus, while the $200 tier stays in place above it with higher overall limits and the company’s exclusive Pro model. (community.openai.com) OpenAI also added a launch promo to make the middle tier feel bigger on day one. Through May 31, 2026, the company says $100 Pro subscribers can get up to 10 times the Codex usage of ChatGPT Plus. (community.openai.com) That change follows another pricing move from just a week earlier. On April 2, OpenAI updated Codex billing for Plus, Pro, Business, and new Enterprise plans so usage aligns with application programming interface token usage instead of per-message pricing. (help.openai.com) So this is not just a cheaper subscription. It is OpenAI rebuilding the ladder for people who code with ChatGPT often enough that “how many tasks can I run” matters more than “do I have access at all.” (community.openai.com; help.openai.com) OpenAI made a second change the same week that most users will never see in a settings menu. Its model release notes say paid users who hit the limit on GPT-5.4 Thinking now fall back to GPT-5.4 mini, so the chat keeps going with a smaller reasoning model instead of stopping cold. (help.openai.com) A separate release note shows the same pattern on the faster side of the product. OpenAI says GPT-5.3 Instant Mini replaced GPT-5 Instant Mini as the hidden fallback after users hit rate limits on GPT-5.3 Instant, and the company says the newer mini model has stronger writing and better contextual awareness. (releasebot.io) That matters because rate limits used to feel like hitting a wall. Now OpenAI is turning them into a lane change: heavy users can pay $100 to push the wall farther out, and when they still hit it, they are more likely to land on a backup model that feels usable. (community.openai.com; help.openai.com; releasebot.io) The competitive target is not hard to spot. Multiple reports tied the move to Anthropic’s Claude Code and Claude Max pricing, with OpenAI trying to give developers a middle option instead of forcing them to choose between a casual $20 plan and a power-user $200 plan. (economictimes.indiatimes.com; 9to5mac.com) The clearest signal is who OpenAI thinks its customers are now. Instead of sorting people into broad buckets like “consumer” and “professional,” it is pricing around workflow intensity: occasional chat, regular coding, and all-day coding with the highest limits. (community.openai.com; openai.com)
