Cmd. /fast
Turn Fast mode on or off, or report its status, to trade reasoning depth for lower latency without switching the active model family.
Key takeaways
/fastturns Fast mode on or off, or reports the current Fast mode status, without changing the model family.- Use
/fast onfor latency-sensitive implementation cleanup and/fast offor/fast statusbefore sensitive reviews. - The persisted
/fasttoggle survives within a session, and the Fast state can be surfaced in the TUI footer via/statusline. - Keep Standard mode for complex design, risky changes, and high-stakes reviews.
Original Command Declaration
/fastOfficial Summary
/fast turns Fast mode on or off, or reports the current Fast mode status.
Usage
/fastGood Examples
- Use
/fast onfor implementation cleanup where latency matters more than depth. - Use
/fast offor/fast statusbefore a sensitive review or architecture turn.
Similar Commands
| Command | Difference | Choose It When |
|---|---|---|
/fast | Changes Fast/Standard service behavior | You want speed control without changing model family |
/model | Changes the active model | You need a different model |
/statusline | Displays footer status | You want Fast state visible all the time |
Use Cases
- Keeping daily implementation fast while reserving standard mode for reviews.
- Checking whether a resumed thread inherited the expected speed mode.
Recent Changes
- The persisted
/fasttoggle made Fast/Standard switching easier to manage during a session. - Fast mode can be shown in the TUI footer through
/statusline.
Cautions
- Use Standard mode for complex design, risky changes, and high-stakes reviews.
- Document when the team expects Fast mode to be allowed.
Release Basis
- Judgment: confirmed addition
- First evidence version:
rust-v0.110.0 - Date (UTC): 2026-03-05
- Release link: https://github.com/openai/codex/releases/tag/rust-v0.110.0
Sources
- Codex slash commands: https://developers.openai.com/codex/cli/slash-commands
- Codex releases: https://github.com/openai/codex/releases