OpenAI reshapes product tiers

OpenAI is redrawing ChatGPT’s product map around GPT-5 and Codex, with new defaults for consumers and new controls for workplace buyers. (help.openai.com) In ChatGPT, GPT-5.3 Instant is now the default for logged-in users, and the company says the older GPT-4o, GPT-4.1, GPT-4.1 mini, OpenAI o4-mini, and GPT-5 Instant and Thinking models were retired from ChatGPT on February 13, 2026. For harder prompts, the “Instant” setting can automatically route a request to GPT-5.4 Thinking. (help.openai.com) On April 9, OpenAI also swapped in GPT-5.3 Instant Mini as the fallback model after users hit GPT-5.3 Instant rate limits. It does not appear in the model picker, and OpenAI says it replaced GPT-5 Instant Mini across ChatGPT and Enterprise and Edu workspaces. (help.openai.com, help.openai.com) The pricing changes are just as important as the model changes. OpenAI’s April 9 release notes show a new $100-a-month Pro tier, while the existing $200-a-month Pro plan stays in place, and Plus remains $20 a month with lower Codex-heavy usage than before. (help.openai.com) For businesses, OpenAI changed the seat structure on April 2. ChatGPT Business and ChatGPT Enterprise now have two seat types — a standard ChatGPT seat and a Codex-only seat — and ChatGPT Business standard seats were cut by $5 to $25 a month on monthly billing or $20 a month on annual billing. (help.openai.com, help.openai.com) OpenAI is also separating how workplace connections are managed. In the Help Center, the company says “connectors” were renamed “apps” in December 2025, and Business release notes on April 20 say workspace owners and admins can now decide whether an app allows all actions, read-only actions, or custom actions, including whether newly added actions stay disabled until review. (help.openai.com, help.openai.com) Enterprise and Edu buyers are getting the same governance push. Their release notes say all sync apps now support in-region data residency in Japan, and admins can now hide SCIM-managed groups from sharing flows for projects and GPTs to limit accidental oversharing. (help.openai.com) Codex is becoming the other half of the bundle. OpenAI’s Codex rate card says pricing now differs across Plus, Pro, Business, and Enterprise and Edu plans, and Enterprise customers are being migrated from older per-message billing toward token-based pricing billed in credits per million tokens. (help.openai.com, help.openai.com) That shift showed up in product form on April 22, when OpenAI introduced “workspace agents” in ChatGPT. OpenAI says these Codex-powered agents are shared across an organization, run in the cloud, keep working when a user is away, and can be used in ChatGPT or Slack under company-set permissions and controls. (openai.com) OpenAI has not formally announced GPT-5.5, but reports on April 23 said the company posted a cryptic “NS41” teaser on X that users interpreted as “5.5.” If that lands soon, it would extend a two-month run in which OpenAI has retired older ChatGPT models, made GPT-5.3 the default, and tied more of the paid experience to team workspaces, admin settings, and Codex usage. (9to5mac.com, help.openai.com)