Yes. OpenAI Codex can remember context across sessions through Memories, an optional feature that is off by default, and it always loads the instructions you keep in AGENTS.md files.
As of November 11, 2026, OpenAI’s Codex Memories documentation says local Codex clients, including the CLI, the IDE extension and the ChatGPT desktop app, keep their own memory store under ~/.codex, separate from ChatGPT web memory. Once enabled, Codex turns useful context from eligible past chats into local memory files in the background, skips active or short sessions and redacts secrets. Turn it on in Settings > Personalization > Enable Codex memories, or add memories = true under [features] in config.toml. In a session, the /memories command controls whether the current chat uses or contributes memories.
For rules that must always apply, OpenAI recommends AGENTS.md. Codex reads a global ~/.codex/AGENTS.md plus project files from the repository root down to your working directory, up to 32 KiB combined by default. Delete Codex memories resets the local store.
Related: Codex CLI 0.167: Power-User Guide to the Agent Command Center, What is the Codex context window? and Does Claude Code have memory?
Source: OpenAI Developers, Codex Memories; OpenAI Developers, Custom instructions with AGENTS.md
Image: ChatGPT logo by OpenAI, public domain (PD shape), via Wikimedia Commons, placed on a white background and resized. Codex and ChatGPT are trademarks of OpenAI. FreeAnswers is not affiliated with or endorsed by OpenAI.