PANews, citing OpenAI Developers, said that OpenAI API supports 12-hour prompt caching. PANews said Tibo added that Codex and Work cache for logged-in users will be retained permanently.
OpenAI Codex Lead Tibo said that over the next 28 days, one clear improvement will be delivered each day to most Codex/Work users; otherwise, a “full reset” will be carried out. The team is currently focused on feature simplification, usage efficiency, breakthrough features, and new models.
