OpenAI has made Codex's full 1 million-token context window available to ChatGPT Plus, Pro, Business, and Enterprise subscribers. Previously, the feature was limited to API-key access.
Users can enable the larger window by adding these settings to
~/.codex/config.toml:model_context_window = 1000000
model_auto_compact_token_limit = 900000
The same options can also be provided through command-line flags.