Codex CLI itself is open source and free to install. You'll need either an OpenAI account with API credits or a ChatGPT subscription. Costs depend on the model you use and how many tokens you consume. With a ChatGPT Plus or Pro subscription, Codex is included within fair-use limits.

What's the difference between Codex CLI and the Codex App?

Codex CLI is the terminal tool you start with codex. It runs entirely in your shell and gives you the full set of commands, slash commands, and flags. The Codex App is a desktop application (macOS and Windows) with a graphical interface. The app has fewer slash commands but adds its own shortcuts, like Cmd+Shift+P for the command palette and Cmd+B for the sidebar. Both use the same models and the same configuration.

Which model should I use with Codex?

As of July 9, 2026, GPT-5.6 Sol is the strongest model available for complex coding tasks, while the previous flagship gpt-5.5 remains usable as a cheaper option. For simpler tasks, gpt-4.1 or gpt-4.1-mini does the job and saves on cost. Switch models any time with /model in the TUI, or with -m at startup.

What's the difference between codex exec and interactive codex?

codex starts an interactive TUI session where you can chat back and forth with the agent. codex exec runs a task non-interactively: it takes a prompt, works through it, and prints the result to stdout. codex exec is ideal for scripts, CI/CD pipelines, and automation.

How do I resume a previous session?

Inside the interactive TUI you type /resume, or codex resume from the terminal. For non-interactive use, run codex exec resume --last followed by your new prompt. Codex stores sessions locally under ~/.codex/sessions/. Sessions started with --ephemeral aren't saved.

What does the sandbox actually do?

The sandbox limits what Codex can do on your system. read-only allows reading only (no file changes, no network). workspace-write allows write access in the working directory only (no network). danger-full-access allows everything: reading, writing anywhere, and network access. workspace-write is enough for most tasks.

Can I use Codex CLI in my CI/CD pipeline?

Yes. Use codex exec with -s workspace-write for automated execution without prompts. With --json you get structured JSONL output. Authenticate with an API key: echo $OPENAI_API_KEY | codex login --with-api-key. Add --skip-git-repo-check if you're running outside a Git repo.

How does Codex differ from Claude Code?

Both are terminal-based AI coding agents. Codex uses OpenAI's GPT models, Claude Code uses Anthropic's Claude models. Codex has especially strong cloud integration (Codex Cloud for remote tasks) and a standalone desktop app alongside the CLI. Claude Code is more terminal-first and adds features like native Git worktrees, hooks, and Extended Thinking. For a detailed comparison, see our article Claude Code vs. OpenAI Codex.

Codex Commands: The Complete Reference

You installed Codex, typed codex into your terminal, and wondered: "Okay, now what?"

Same here. OpenAI's coding agent ships with dozens of CLI commands, more than 40 slash commands in the interactive TUI, and a separate desktop app with its own shortcuts. Without a reference, you lose track fast.

This article lays out every Codex command in one place. Each CLI command, every slash command, all flags, every keyboard shortcut for CLI and App. It's the most thorough Codex reference you'll find in English. If you'd rather compare Codex head-to-head with Claude Code, our in-depth Claude Code vs. OpenAI Codex comparison has you covered.

TL;DRKey Takeaways

30+ CLI commands cover the whole tool, from codex exec (non-interactive execution) to codex review (code review) and codex cloud (cloud tasks)
Over 40 slash commands in the composer control the model, permissions, sessions, plan mode, and more, with parameter and (where documented) introduction-version columns
Three sandbox modes (read-only, workspace-write, danger-full-access) and four approval policies control what Codex can do on your system
Codex Cloud offloads long-running tasks to remote sandboxed environments, with best-of-N attempts on demand

Installing and Starting Codex

Codex CLI installs via npm or Homebrew. On first launch, Codex asks you to sign in with your ChatGPT account or an API key.

# Install via npm
npm install -g @openai/codex

# Or via Homebrew (native Rust binary, faster)
brew install codex

# Sign in interactively (opens browser for OAuth)
codex login

# Sign in with an API key
echo $OPENAI_API_KEY | codex login --with-api-key

# Check login status
codex login status

# Sign out
codex logout

# Start an interactive TUI session
codex

# Start with an initial prompt
codex "Explain the architecture of this project"

# Run non-interactively (for scripts)
codex exec -s workspace-write -m gpt-5.5 "Refactor auth.ts"

# Launch the desktop app
codex app

Tip

Always start Codex inside the right project folder. You can also set the working directory explicitly with --cd /path/to/project.

CLI Commands (Overview)

Codex comes in two flavors. The CLI runs in your terminal. The App is a desktop application for macOS and Windows. The "Example" column shows the typical usage of each command.

Commandcodex

DescriptionStart an interactive TUI session in the terminal

Examplecodex

Commandcodex app

DescriptionLaunch the Codex desktop app (macOS/Windows)

Examplecodex app

Commandcodex exec

DescriptionRun a task non-interactively, stream output to stdout

Examplecodex exec "Fix bug"

Commandcodex exec resume

DescriptionResume a previous session non-interactively

Examplecodex exec resume --last "Continue task"

Commandcodex review

DescriptionCode review (uncommitted, branch diff, or commit)

Examplecodex review --uncommitted

Commandcodex apply

DescriptionApply a diff from a Codex session or cloud task

Examplecodex apply TASK_ID

Commandcodex resume

DescriptionResume an interactive session (session picker)

Examplecodex resume

Commandcodex fork

DescriptionFork a session into a new thread

Examplecodex fork

Commandcodex cloud exec

DescriptionSubmit a cloud task in a sandbox environment

Examplecodex cloud exec --env ENV_ID "..."

Commandcodex cloud list

DescriptionList cloud tasks (with filter and pagination)

Examplecodex cloud list

Commandcodex cloud status

DescriptionShow the status of a cloud task

Examplecodex cloud status ID

Commandcodex cloud diff

DescriptionShow the diff of a cloud task

Examplecodex cloud diff ID

Commandcodex cloud apply

DescriptionApply a cloud task diff locally

Examplecodex cloud apply ID

Commandcodex mcp add

DescriptionAdd an MCP server (stdio or HTTP)

Examplecodex mcp add name -- cmd

Commandcodex mcp list

DescriptionList configured MCP servers

Examplecodex mcp list

Commandcodex mcp get

DescriptionShow an MCP server's configuration

Examplecodex mcp get name

Commandcodex mcp remove

DescriptionRemove an MCP server

Examplecodex mcp remove name

Commandcodex mcp login

DescriptionOAuth login for an MCP server

Examplecodex mcp login name

Commandcodex mcp logout

DescriptionRemove an MCP server's OAuth credentials

Examplecodex mcp logout name

Commandcodex features list

DescriptionShow feature flags with status and maturity

Examplecodex features list

Commandcodex features enable/disable

DescriptionEnable or disable a feature flag persistently

Examplecodex features enable NAME

Commandcodex login

DescriptionSign in via OAuth or API key

Examplecodex login

Commandcodex login status

DescriptionShow the current login status

Examplecodex login status

Commandcodex logout

DescriptionRemove stored credentials

Examplecodex logout

Commandcodex sandbox

DescriptionRun a command in the Codex sandbox (macOS/Linux/Windows)

Examplecodex sandbox -- cmd

Commandcodex completion

DescriptionGenerate shell completions (bash, zsh, fish, elvish, PowerShell)

Examplecodex completion zsh

Commandcodex update

DescriptionCheck for updates and install the new version

Examplecodex update

Commandcodex execpolicy

DescriptionCheck exec-policy rule files before saving

Examplecodex execpolicy check --rules rules.toml -- npm test

Commandcodex mcp-server

DescriptionRun Codex itself as an MCP server (stdio transport)

Examplecodex mcp-server

Commandcodex plugin marketplace

DescriptionManage plugin marketplace sources (add, remove, upgrade)

Examplecodex plugin marketplace add URL

Commandcodex remote-control

DescriptionStart a headless app server, controllable from the Codex app or Chrome extension (since 0.130)

Examplecodex remote-control

	Description	Example
`codex`	Start an interactive TUI session in the terminal	`codex`
`codex app`	Launch the Codex desktop app (macOS/Windows)	`codex app`
`codex exec`	Run a task non-interactively, stream output to stdout	`codex exec "Fix bug"`
`codex exec resume`	Resume a previous session non-interactively	`codex exec resume --last "Continue task"`
`codex review`	Code review (uncommitted, branch diff, or commit)	`codex review --uncommitted`
`codex apply`	Apply a diff from a Codex session or cloud task	`codex apply TASK_ID`
`codex resume`	Resume an interactive session (session picker)	`codex resume`
`codex fork`	Fork a session into a new thread	`codex fork`
`codex cloud exec`	Submit a cloud task in a sandbox environment	`codex cloud exec --env ENV_ID "..."`
`codex cloud list`	List cloud tasks (with filter and pagination)	`codex cloud list`
`codex cloud status`	Show the status of a cloud task	`codex cloud status ID`
`codex cloud diff`	Show the diff of a cloud task	`codex cloud diff ID`
`codex cloud apply`	Apply a cloud task diff locally	`codex cloud apply ID`
`codex mcp add`	Add an MCP server (stdio or HTTP)	`codex mcp add name -- cmd`
`codex mcp list`	List configured MCP servers	`codex mcp list`
`codex mcp get`	Show an MCP server's configuration	`codex mcp get name`
`codex mcp remove`	Remove an MCP server	`codex mcp remove name`
`codex mcp login`	OAuth login for an MCP server	`codex mcp login name`
`codex mcp logout`	Remove an MCP server's OAuth credentials	`codex mcp logout name`
`codex features list`	Show feature flags with status and maturity	`codex features list`
`codex features enable/disable`	Enable or disable a feature flag persistently	`codex features enable NAME`
`codex login`	Sign in via OAuth or API key	`codex login`
`codex login status`	Show the current login status	`codex login status`
`codex logout`	Remove stored credentials	`codex logout`
`codex sandbox`	Run a command in the Codex sandbox (macOS/Linux/Windows)	`codex sandbox -- cmd`
`codex completion`	Generate shell completions (bash, zsh, fish, elvish, PowerShell)	`codex completion zsh`
`codex update`	Check for updates and install the new version	`codex update`
`codex execpolicy`	Check exec-policy rule files before saving	`codex execpolicy check --rules rules.toml -- npm test`
`codex mcp-server`	Run Codex itself as an MCP server (stdio transport)	`codex mcp-server`
`codex plugin marketplace`	Manage plugin marketplace sources (add, remove, upgrade)	`codex plugin marketplace add URL`
`codex remote-control`	Start a headless app server, controllable from the Codex app or Chrome extension (since 0.130)	`codex remote-control`

Note

codex exec is the command you'll use most for automation. It runs tasks non-interactively and prints the result to stdout. Perfect for scripts and CI/CD pipelines.

codex exec: Flags for Non-Interactive Execution

codex exec is your workhorse for automated workflows. Without any flags, Codex waits for human approval, and scripts hang. So always set a sandbox mode.

# Standard pattern for automated execution
codex exec -s workspace-write -m gpt-5.5 "your prompt"

# Maximum reasoning depth
codex exec -s workspace-write -m gpt-5.5 \
  -c model_reasoning_effort='"xhigh"' "complex task"

# Read-only for analysis (no file changes)
codex exec -s read-only -m gpt-5.5 "Analyze the architecture"

Flag--model, -m

DescriptionOverrides the model from your config

Example-m gpt-5.5

Flag--sandbox, -s

DescriptionSandbox mode: read-only, workspace-write, or danger-full-access

Example-s workspace-write

Flag--ask-for-approval, -a

DescriptionApproval policy: untrusted, on-failure, on-request, never

Example-a on-request

Flag--cd, -C

DescriptionSets the working directory for the agent

Example-C /path/to/project

Flag--add-dir

DescriptionGrants additional writable directories

Example--add-dir ../backend

Flag--image, -i

DescriptionAttaches images to the prompt (repeatable)

Example-i screenshot.png

Flag--config, -c

DescriptionOverrides config.toml values (TOML syntax)

Example-c model_reasoning_effort='"xhigh"'

Flag--profile, -p

DescriptionLoads a named config profile from config.toml

Example-p fast-profile

Flag--enable / --disable

DescriptionEnables or disables a feature flag for this run

Example--enable search_tool

Flag--json

DescriptionOutputs events as JSONL (one JSON line per event)

Examplecodex exec --json "..."

Flag--output-last-message, -o

DescriptionSaves the agent's last message to a file

Example-o /tmp/result.txt

Flag--output-schema

DescriptionJSON Schema file that constrains the final response

Example--output-schema schema.json

Flag--ephemeral

DescriptionSession is not persisted to disk

Examplecodex exec --ephemeral "..."

Flag--skip-git-repo-check

DescriptionAllows running outside a Git repo

Examplecodex exec --skip-git-repo-check "..."

Flag--color

DescriptionControls ANSI colors: always, never, auto

Example--color never

Flag--oss

DescriptionUses the local open-source provider (Ollama)

Examplecodex exec --oss "..."

Flag--full-auto

DescriptionDeprecated. Alias for -a on-request -s workspace-write. Use -s workspace-write directly instead

Examplecodex exec --full-auto "..."

Flag	Description	Example
`--model, -m`	Overrides the model from your config	`-m gpt-5.5`
`--sandbox, -s`	Sandbox mode: read-only, workspace-write, or danger-full-access	`-s workspace-write`
`--ask-for-approval, -a`	Approval policy: untrusted, on-failure, on-request, never	`-a on-request`
`--cd, -C`	Sets the working directory for the agent	`-C /path/to/project`
`--add-dir`	Grants additional writable directories	`--add-dir ../backend`
`--image, -i`	Attaches images to the prompt (repeatable)	`-i screenshot.png`
`--config, -c`	Overrides config.toml values (TOML syntax)	`-c model_reasoning_effort='"xhigh"'`
`--profile, -p`	Loads a named config profile from config.toml	`-p fast-profile`
`--enable / --disable`	Enables or disables a feature flag for this run	`--enable search_tool`
`--json`	Outputs events as JSONL (one JSON line per event)	`codex exec --json "..."`
`--output-last-message, -o`	Saves the agent's last message to a file	`-o /tmp/result.txt`
`--output-schema`	JSON Schema file that constrains the final response	`--output-schema schema.json`
`--ephemeral`	Session is not persisted to disk	`codex exec --ephemeral "..."`
`--skip-git-repo-check`	Allows running outside a Git repo	`codex exec --skip-git-repo-check "..."`
`--color`	Controls ANSI colors: always, never, auto	`--color never`
`--oss`	Uses the local open-source provider (Ollama)	`codex exec --oss "..."`
`--full-auto`	Deprecated. Alias for `-a on-request -s workspace-write`. Use `-s workspace-write` directly instead	`codex exec --full-auto "..."`

Warning

--full-auto is deprecated. Use -s workspace-write instead (or -s workspace-write -a on-request for the same behavior).

Common Patterns

# Attach images (screenshots, design specs)
codex exec -s workspace-write -m gpt-5.5 -i screenshot.png "Implement this UI"

# Multiple writable directories
codex exec -s workspace-write -C apps/frontend --add-dir ../backend "Sync API types"

# JSON output for scripts
codex exec -s workspace-write --json "List all TODO comments"

# Save the result to a file
codex exec -s workspace-write -o /tmp/result.txt "Summarize this codebase"

# Structured output with JSON Schema
codex exec -s workspace-write --output-schema schema.json "Extract all API endpoints"

# Pipe the prompt via stdin
cat prompt.txt | codex exec -s workspace-write -

# Use a config profile
codex exec -s workspace-write -p thorough "Thorough analysis"

# Local open-source model (Ollama)
codex exec --oss -s workspace-write "Explain this function"

Code Review

codex review analyzes code changes non-interactively and prints the results to stdout. It doesn't modify files.

Flag--uncommitted

DescriptionReviews staged, unstaged, and untracked changes

Examplecodex review --uncommitted

Flag--base

DescriptionReviews changes against a base branch

Examplecodex review --base main

Flag--commit

DescriptionReviews a specific commit

Examplecodex review --commit abc123

Flag--title

DescriptionOptional title for the review summary

Examplecodex review --title "Add auth"

FlagPROMPT

DescriptionCustom review instructions (can also be piped via stdin with -)

Examplecodex review "Focus on security"

Flag	Description	Example
`--uncommitted`	Reviews staged, unstaged, and untracked changes	`codex review --uncommitted`
`--base`	Reviews changes against a base branch	`codex review --base main`
`--commit`	Reviews a specific commit	`codex review --commit abc123`
`--title`	Optional title for the review summary	`codex review --title "Add auth"`
`PROMPT`	Custom review instructions (can also be piped via stdin with -)	`codex review "Focus on security"`

# Review uncommitted changes (staged + unstaged + untracked)
codex review --uncommitted

# Review changes against a branch (before opening a PR)
codex review --base main

# Review a specific commit
codex review --commit abc123

# Review with a title and a custom focus
codex review --base main --title "Add auth" "Focus on security"

# Review via exec with JSON output
codex exec review --uncommitted --json

Session Management

Codex stores sessions locally under ~/.codex/sessions/. You can resume them any time, fork them, or pick them up non-interactively.

# Interactive: open the session picker
codex resume

# Interactive: resume the most recent session directly
codex resume --last

# Non-interactive: resume the latest session with a new prompt
codex exec resume --last -s workspace-write -m gpt-5.5 \
  "Now add error handling"

# Resume a specific session by UUID
codex exec resume 7f9f9a2e-1b3c-4c7a... -s workspace-write -m gpt-5.5 \
  "Implement the plan"

# Fork a session (new thread from existing context)
codex fork --last

# Apply a diff from a local session
codex apply TASK_ID

# Apply a diff from a cloud task
codex cloud apply TASK_ID

# Apply a specific attempt (when using --attempts N)
codex cloud apply TASK_ID --attempt 2

Tip

Sessions started with --ephemeral aren't saved to disk. Use this for quick one-off questions where you don't need any history.

Codex Cloud

Codex Cloud runs tasks in sandboxed remote environments. All subcommands are non-interactive. The --env flag is required.

Flag--env

DescriptionRequired: target environment ID for the cloud task

Example--env ENV_ID

Flag--attempts

DescriptionNumber of attempts (best-of-N), default: 1, max: 4

Example--attempts 3

Flag--branch

DescriptionGit branch the task should run on

Example--branch feature/auth

Flag--json

DescriptionMachine-readable JSON output (for codex cloud list)

Examplecodex cloud list --json

Flag--limit

DescriptionMaximum number of tasks returned (1-20)

Examplecodex cloud list --limit 5

Flag--cursor

DescriptionPagination cursor from the previous call

Examplecodex cloud list --cursor CURSOR

Flag--attempt

DescriptionAttempt number for diff/apply (when using --attempts N)

Examplecodex cloud diff TASK_ID --attempt 2

Flag	Description	Example
`--env`	Required: target environment ID for the cloud task	`--env ENV_ID`
`--attempts`	Number of attempts (best-of-N), default: 1, max: 4	`--attempts 3`
`--branch`	Git branch the task should run on	`--branch feature/auth`
`--json`	Machine-readable JSON output (for codex cloud list)	`codex cloud list --json`
`--limit`	Maximum number of tasks returned (1-20)	`codex cloud list --limit 5`
`--cursor`	Pagination cursor from the previous call	`codex cloud list --cursor CURSOR`
`--attempt`	Attempt number for diff/apply (when using --attempts N)	`codex cloud diff TASK_ID --attempt 2`

# Submit a task
codex cloud exec --env ENV_ID "Refactor the payment module"

# Best-of-3: run three attempts
codex cloud exec --env ENV_ID --attempts 3 "Fix the flaky test"

# Run on a specific branch
codex cloud exec --env ENV_ID --branch feature/auth "Add OAuth"

# List tasks
codex cloud list
codex cloud list --json --limit 5

# Check the status
codex cloud status TASK_ID

# Show the diff (specific attempt)
codex cloud diff TASK_ID --attempt 2

Sandbox Modes

The sandbox limits what Codex can do on your system. You pick the mode with -s or --sandbox.

Moderead-only

ReadAnywhere

WriteNowhere

NetworkNo

When to useQuestions, explanations, code analysis

Modeworkspace-write

ReadAnywhere

WriteWorking dir only

NetworkNo

When to useDefault for most tasks

Modedanger-full-access

ReadAnywhere

WriteAnywhere

NetworkYes

When to useOnly when truly needed (e.g., installing packages)

Mode	Read	Write	Network	When to use
`read-only`	Anywhere	Nowhere	No	Questions, explanations, code analysis
`workspace-write`	Anywhere	Working dir only	No	Default for most tasks
`danger-full-access`	Anywhere	Anywhere	Yes	Only when truly needed (e.g., installing packages)

Codex also ships with a codex sandbox command that lets you run any command inside the sandbox without involving the agent:

# Run a command in the sandbox (macOS)
codex sandbox macos -- npm test

# Log denied accesses for debugging
codex sandbox macos --log-denials -- ./build.sh

# Linux (Landlock + seccomp)
codex sandbox linux -- python3 script.py

Approval Policies

Approval policies control when Codex asks for permission before running a command. You set them with -a or --ask-for-approval.

Policyuntrusted

BehaviorOnly trusted commands (ls, cat, sed) run without approval

Policyon-failure

BehaviorAll commands run; approval only on failure

Policyon-request

BehaviorThe model decides when to ask for approval

Policynever

BehaviorNever asks for approval; failures go straight back to the model

Policy	Behavior
`untrusted`	Only trusted commands (ls, cat, sed) run without approval
`on-failure`	All commands run; approval only on failure
`on-request`	The model decides when to ask for approval
`never`	Never asks for approval; failures go straight back to the model

Note

Inside the interactive TUI you switch approval mode with /permissions. Three presets are available: Auto (default), Read Only, and Full Access.

MCP Servers (Model Context Protocol)

Codex can connect to MCP servers for extra tools (databases, APIs, GitHub, and more). All management commands are non-interactive.

# List configured servers
codex mcp list
codex mcp list --json

# Show a server's config
codex mcp get my-server
codex mcp get my-server --json

# Add a stdio server
codex mcp add my-server -- npx -y @my/mcp-server
codex mcp add my-server --env API_KEY=sk-123 -- node server.js

# Add an HTTP server
codex mcp add my-server --url https://mcp.example.com/sse
codex mcp add my-server --url https://mcp.example.com \
  --bearer-token-env-var MY_TOKEN_VAR

# Remove a server
codex mcp remove my-server

# OAuth login for an MCP server
codex mcp login my-server --scopes "read,write"
codex mcp logout my-server

Tip

With codex mcp-server you can run Codex itself as an MCP server. That makes it possible to plug Codex into other tools that speak MCP.

Slash Commands (CLI and App)

Slash commands are available in the interactive TUI (started with codex) and in the desktop app. Type / in the composer to open the list. While a task is running, you can type a slash command and press Tab to queue it for the next turn.

The CLI TUI has far more slash commands than the app. The "Since" column shows the Codex version a command is documented from in the official release notes (where available, otherwise "n/a").

Command/agent

Parameter

Since0.95

DescriptionSwitch the active agent thread

Command/apps

Parameter

Since0.93

DescriptionBrowse connectors and apps and insert them into the prompt

Command/approve

Parameter

Since0.129

DescriptionApprove one retry after a denied auto review

Command/archive

Parameter

Since0.136

DescriptionArchive the current session and exit Codex

Command/clear

Parameter

Since0.105

DescriptionClear the terminal and start a fresh chat

Command/compact

Parameter

Since0.11

DescriptionSummarize the visible conversation to free tokens

Command/copy

Parameter

Since0.105

DescriptionCopy the latest completed Codex output (also Ctrl+O)

Command/debug-config

Parameter

Since0.96

DescriptionPrint config layers and diagnostics

Command/delete

Parameter

Since0.140

DescriptionPermanently delete the current session and exit Codex

Command/diff

Parameter

Since0.23

DescriptionShow the Git diff, including untracked files

Command/exit

Parameter

Since0.53

DescriptionExit the CLI (alias /quit)

Command/experimental

Parameter

Since0.74

DescriptionToggle experimental features on or off

Command/fast

Parameteron | off | status

Since0.110

DescriptionToggle the Fast service tier

Command/feedback

Parameter

Since0.50

DescriptionSend logs to the Codex maintainers (also App)

Command/fork

Parameter

Since0.88

DescriptionFork the current conversation into a new thread

Command/goal

Parametergoal | pause | resume | clear

Since0.128

DescriptionSet, pause, or clear a task goal

Command/hooks

Parameter

Since0.129

DescriptionView and manage lifecycle hooks

Command/ide

Parametertext

Since0.129

DescriptionInclude open files, selection, and IDE context

Command/import

Parameter

Since0.140

DescriptionImport Claude Code setup, project files, and chats

Command/init

Parameter

Since0.28

DescriptionGenerate an AGENTS.md scaffold as project context

Command/keymap

Parameter

Since0.129

DescriptionRemap the TUI keyboard shortcuts

Command/logout

Parameter

Sincen/a

DescriptionSign out of Codex

Command/mcp

Parameterverbose

Since0.23

DescriptionList the configured MCP tools (also App)

Command/memories

Parameter

Sincen/a

DescriptionConfigure memory use and generation

Command/mention

Parameterpath

Sincen/a

DescriptionAttach a file to the conversation

Command/model

Parametermodel

Since0.23

DescriptionChoose the active model and reasoning effort

Command/new

Parameter

Since0.59

DescriptionStart a new conversation in the same CLI session

Command/permissions

Parameter

Since0.89

DescriptionSet what Codex can do without asking first

Command/personality

Parameterstyle

Since0.92

DescriptionChoose a communication style for responses

Command/plan

Parameterprompt

Since0.93

DescriptionSwitch to plan mode (optional inline prompt)

Command/plugins

Parameter

Since0.117

DescriptionBrowse installed and discoverable plugins

Command/ps

Parameter

Since0.76

DescriptionShow experimental background terminals

Command/raw

Parameteron | off

Since0.129

DescriptionToggle raw scrollback mode

Command/resume

Parameter

Since0.65

DescriptionResume a saved session via the session picker

Command/review

Parameter

Since0.39

DescriptionAsk Codex to review your working tree (also App)

Command/sandbox-add-read-dir

Parameterdirectory

Sincen/a

DescriptionGrant the sandbox read access to a directory (Windows only)

Command/side

Parametertext

Since0.122

DescriptionStart an ephemeral side conversation (alias /btw)

Command/skills

Parameter

Since0.65

DescriptionBrowse and use skills

Command/status

Parameter

Since0.23

DescriptionShow model, approval policy, writable paths, and token usage (also App)

Command/statusline

Parameter

Since0.99

DescriptionConfigure the TUI status-line fields

Command/stop

Parameter

Since0.115

DescriptionStop all background terminals

Command/theme

Parameter

Since0.105

DescriptionChoose a syntax-highlighting theme

Command/title

Parameter

Since0.117

DescriptionConfigure the terminal window or tab title

Command/usage

Parameterdaily | weekly | cumulative

Since0.140

DescriptionView token usage or redeem a rate-limit reset

Command/vim

Parameter

Since0.129

DescriptionToggle Vim mode for the composer

	Parameter		Description
`/agent`		0.95	Switch the active agent thread
`/apps`		0.93	Browse connectors and apps and insert them into the prompt
`/approve`		0.129	Approve one retry after a denied auto review
`/archive`		0.136	Archive the current session and exit Codex
`/clear`		0.105	Clear the terminal and start a fresh chat
`/compact`		0.11	Summarize the visible conversation to free tokens
`/copy`		0.105	Copy the latest completed Codex output (also `Ctrl+O`)
`/debug-config`		0.96	Print config layers and diagnostics
`/delete`		0.140	Permanently delete the current session and exit Codex
`/diff`		0.23	Show the Git diff, including untracked files
`/exit`		0.53	Exit the CLI (alias `/quit`)
`/experimental`		0.74	Toggle experimental features on or off
`/fast`	`on \| off \| status`	0.110	Toggle the Fast service tier
`/feedback`		0.50	Send logs to the Codex maintainers (also App)
`/fork`		0.88	Fork the current conversation into a new thread
`/goal`	`goal \| pause \| resume \| clear`	0.128	Set, pause, or clear a task goal
`/hooks`		0.129	View and manage lifecycle hooks
`/ide`	text	0.129	Include open files, selection, and IDE context
`/import`		0.140	Import Claude Code setup, project files, and chats
`/init`		0.28	Generate an AGENTS.md scaffold as project context
`/keymap`		0.129	Remap the TUI keyboard shortcuts
`/logout`		n/a	Sign out of Codex
`/mcp`	`verbose`	0.23	List the configured MCP tools (also App)
`/memories`		n/a	Configure memory use and generation
`/mention`	path	n/a	Attach a file to the conversation
`/model`	model	0.23	Choose the active model and reasoning effort
`/new`		0.59	Start a new conversation in the same CLI session
`/permissions`		0.89	Set what Codex can do without asking first
`/personality`	style	0.92	Choose a communication style for responses
`/plan`	prompt	0.93	Switch to plan mode (optional inline prompt)
`/plugins`		0.117	Browse installed and discoverable plugins
`/ps`		0.76	Show experimental background terminals
`/raw`	`on \| off`	0.129	Toggle raw scrollback mode
`/resume`		0.65	Resume a saved session via the session picker
`/review`		0.39	Ask Codex to review your working tree (also App)
`/sandbox-add-read-dir`	directory	n/a	Grant the sandbox read access to a directory (Windows only)
`/side`	text	0.122	Start an ephemeral side conversation (alias `/btw`)
`/skills`		0.65	Browse and use skills
`/status`		0.23	Show model, approval policy, writable paths, and token usage (also App)
`/statusline`		0.99	Configure the TUI status-line fields
`/stop`		0.115	Stop all background terminals
`/theme`		0.105	Choose a syntax-highlighting theme
`/title`		0.117	Configure the terminal window or tab title
`/usage`	`daily \| weekly \| cumulative`	0.140	View token usage or redeem a rate-limit reset
`/vim`		0.129	Toggle Vim mode for the composer

Keyboard Shortcuts: CLI (TUI)

These shortcuts work in the interactive terminal interface (started with codex). You can remap them with /keymap.

ShortcutCtrl+C

DescriptionCancels the current execution

ContextWhile the agent is working

ShortcutCtrl+D

DescriptionCloses the CLI

ContextWhen the input buffer is empty

ShortcutCtrl+G

DescriptionOpens external editor (VISUAL or EDITOR)

ContextFor long prompts

ShortcutCtrl+L

DescriptionClears the screen (without starting a new chat)

ContextWhen the terminal gets too cluttered

ShortcutCtrl+O

DescriptionCopies the last response to the clipboard

ContextAfter a response

ShortcutCtrl+R

DescriptionReverse search through prompt history

ContextIn the composer

ShortcutAlt+,

DescriptionLowers the reasoning depth

ContextDuring the session

ShortcutAlt+.

DescriptionRaises the reasoning depth

ContextDuring the session

ShortcutUp / Down

DescriptionNavigates the draft history in the composer

ContextIn the input field

ShortcutTab

DescriptionQueues follow-up text, slash commands, or shell commands

ContextWhile the agent is working

ShortcutEsc + Esc

DescriptionEdits the previous user message

ContextIn the composer

Shortcut@

DescriptionFuzzy file search for attaching files

ContextIn the input field

Shortcut!

DescriptionRuns a shell command directly

ContextAt the start of the input

Shortcut	Description	Context
`Ctrl+C`	Cancels the current execution	While the agent is working
`Ctrl+D`	Closes the CLI	When the input buffer is empty
`Ctrl+G`	Opens external editor (VISUAL or EDITOR)	For long prompts
`Ctrl+L`	Clears the screen (without starting a new chat)	When the terminal gets too cluttered
`Ctrl+O`	Copies the last response to the clipboard	After a response
`Ctrl+R`	Reverse search through prompt history	In the composer
`Alt+,`	Lowers the reasoning depth	During the session
`Alt+.`	Raises the reasoning depth	During the session
`Up / Down`	Navigates the draft history in the composer	In the input field
`Tab`	Queues follow-up text, slash commands, or shell commands	While the agent is working
`Esc + Esc`	Edits the previous user message	In the composer
`@`	Fuzzy file search for attaching files	In the input field
`!`	Runs a shell command directly	At the start of the input

Keyboard Shortcuts: Codex App

The desktop app (started with codex app) has its own shortcuts that follow common IDE conventions. On Windows, replace Cmd with Ctrl.

ShortcutCmd+Shift+P / Cmd+K

DescriptionOpens the command palette

ShortcutCmd+,

DescriptionOpens settings

ShortcutCmd+O

DescriptionOpens a folder

ShortcutCmd+N / Cmd+Shift+O

DescriptionNew thread

ShortcutCmd+F

DescriptionFind in thread

ShortcutCmd+B

DescriptionToggle sidebar

ShortcutCmd+Option+B

DescriptionToggle diff panel

ShortcutCmd+J

DescriptionToggle terminal

ShortcutCmd+Shift+[ / ]

DescriptionSwitch to previous / next thread

ShortcutCmd+[ / ]

DescriptionNavigate back / forward

ShortcutCmd++ / Cmd+-

DescriptionIncrease / decrease font size

ShortcutCtrl+L

DescriptionClear terminal

ShortcutCtrl+M

DescriptionStart dictation

Shortcut	Description
`Cmd+Shift+P / Cmd+K`	Opens the command palette
`Cmd+,`	Opens settings
`Cmd+O`	Opens a folder
`Cmd+N / Cmd+Shift+O`	New thread
`Cmd+F`	Find in thread
`Cmd+B`	Toggle sidebar
`Cmd+Option+B`	Toggle diff panel
`Cmd+J`	Toggle terminal
`Cmd+Shift+[ / ]`	Switch to previous / next thread
`Cmd+[ / ]`	Navigate back / forward
`Cmd++ / Cmd+-`	Increase / decrease font size
`Ctrl+L`	Clear terminal
`Ctrl+M`	Start dictation

Note

The app also supports deeplinks via the codex:// scheme, for example codex://settings for settings or codex://new?prompt=your+prompt for a new thread with a prompt.

Configuration (config.toml)

Codex config lives at ~/.codex/config.toml. You can set defaults and define named profiles.

Settingmodel

DescriptionDefault model for all sessions

Possible values"gpt-5.5" (recommended), "gpt-5.4", "gpt-4.1", "gpt-4.1-mini"

Settingmodel_reasoning_effort

DescriptionReasoning depth for complex tasks

Possible values"xhigh", "high", "medium", "low"

Settingpersonality

DescriptionCommunication style of the agent

Possible values"pragmatic", "friendly", "none"

Settingweb_search

DescriptionWeb search mode

Possible values"cached" (default), "live", "disabled"

Setting	Description	Possible values
`model`	Default model for all sessions	`"gpt-5.5"` (recommended), `"gpt-5.4"`, `"gpt-4.1"`, `"gpt-4.1-mini"`
`model_reasoning_effort`	Reasoning depth for complex tasks	`"xhigh"`, `"high"`, `"medium"`, `"low"`
`personality`	Communication style of the agent	`"pragmatic"`, `"friendly"`, `"none"`
`web_search`	Web search mode	`"cached"` (default), `"live"`, `"disabled"`

# ~/.codex/config.toml

# Default settings
model = "gpt-5.5"
model_reasoning_effort = "high"

# Profile for quick tasks
[profiles.quick]
model = "gpt-4.1-mini"
model_reasoning_effort = "low"

# Profile for thorough work
[profiles.thorough]
model = "gpt-5.5"
model_reasoning_effort = "xhigh"

# Project trust level
[projects."/path/to/project"]
trust_level = "trusted"

# Override a single value
codex exec -s workspace-write -c model='"gpt-5.5"' "Task"

# Set sandbox permissions
codex exec -s workspace-write \
  -c 'sandbox_permissions=["disk-full-read-access"]' "Task"

# Load a profile with -p
codex exec -s workspace-write -p quick "Quick question"

Feature Flags

Feature flags control both experimental and stable features. Changes via enable and disable are written permanently to config.toml. Per-run toggles with --enable and --disable only apply to the current run.

# List all feature flags with status and maturity
codex features list

# Permanently enable / disable a feature
codex features enable search_tool
codex features disable shell_snapshot

# Enable for this run only (not persistent)
codex exec -s workspace-write --enable search_tool \
  --disable shell_snapshot "Task"

Global Flags

These flags are available on almost every Codex command:

Flag--config, -c

DescriptionOverrides config.toml values. Dotted paths for nesting. Values are parsed as TOML.

Example-c model_reasoning_effort='"high"'

Flag--enable

DescriptionEnables a feature flag for this run (repeatable)

Examplecodex --enable search_tool

Flag--disable

DescriptionDisables a feature flag for this run (repeatable)

Examplecodex --disable search_tool

Flag--help, -h

DescriptionShows help

Examplecodex --help

Flag--version, -V

DescriptionShows the version number

Examplecodex --version

Flag	Description	Example
`--config, -c`	Overrides config.toml values. Dotted paths for nesting. Values are parsed as TOML.	`-c model_reasoning_effort='"high"'`
`--enable`	Enables a feature flag for this run (repeatable)	`codex --enable search_tool`
`--disable`	Disables a feature flag for this run (repeatable)	`codex --disable search_tool`
`--help, -h`	Shows help	`codex --help`
`--version, -V`	Shows the version number	`codex --version`

Troubleshooting

ProblemAuthentication fails

Likely causeStale token or wrong API key

SolutionRun codex logout, then codex login again

ProblemCodex hangs on codex exec

Likely causeNo sandbox mode set (waiting for approval)

SolutionAlways pass -s workspace-write or -a on-request

ProblemCommand blocked in sandbox

Likely causeNetwork access or write access outside the working directory

SolutionUse -s danger-full-access (only when needed)

ProblemMCP server doesn't connect

Likely causeMissing dependencies or wrong path

SolutionCheck codex mcp get server-name, then remove and add again

ProblemSession won't resume

Likely causeSession was started with --ephemeral

SolutionRestart without --ephemeral so sessions are saved

ProblemCloud task shows no diff

Likely causeTask is still running or has failed

SolutionCheck codex cloud status TASK_ID, then try again

Problem	Likely cause	Solution
Authentication fails	Stale token or wrong API key	Run codex logout, then codex login again
Codex hangs on codex exec	No sandbox mode set (waiting for approval)	Always pass -s workspace-write or -a on-request
Command blocked in sandbox	Network access or write access outside the working directory	Use -s danger-full-access (only when needed)
MCP server doesn't connect	Missing dependencies or wrong path	Check codex mcp get server-name, then remove and add again
Session won't resume	Session was started with --ephemeral	Restart without --ephemeral so sessions are saved
Cloud task shows no diff	Task is still running or has failed	Check codex cloud status TASK_ID, then try again

# Diagnose the config layers
# (in the interactive TUI)
/debug-config

# Show available models as JSON
codex debug models

# Show only the bundled model catalog (no API refresh)
codex debug models --bundled

# Validate exec policy rules (before saving)
codex execpolicy check --rules rules.toml -- npm test

# Log sandbox denials (debugging)
codex sandbox macos --log-denials -- ./build.sh

Codex Commands: The Complete Reference

Installing and Starting Codex

CLI Commands (Overview)

codex exec: Flags for Non-Interactive Execution

Common Patterns

Code Review

Session Management

Codex Cloud

Sandbox Modes

Approval Policies

MCP Servers (Model Context Protocol)

Slash Commands (CLI and App)

Keyboard Shortcuts: CLI (TUI)

Keyboard Shortcuts: Codex App

Configuration (config.toml)

Feature Flags

Global Flags

Troubleshooting

FAQ

Frequently Asked Questions

Finn Hillebrandt

Similar Articles

Claude Code Statistics 2026: Key Numbers, Data & Facts

OpenAI Codex Statistics 2026: Key Numbers, Data & Facts

How to Use Claude Code for Free (And What the Limits Are)

Claude Code Pricing: How Much Does Claude Code Really Cost?

Claude Code Alternatives 2026: 12 AI Coding Tools Compared

Claude Code: The Complete Beginner's Guide

Installing and Starting Codex

CLI Commands (Overview)

codex exec: Flags for Non-Interactive Execution

Common Patterns

Code Review

Session Management

Codex Cloud

Sandbox Modes

Approval Policies

MCP Servers (Model Context Protocol)

Slash Commands (CLI and App)

Keyboard Shortcuts: CLI (TUI)

Keyboard Shortcuts: Codex App

Configuration (config.toml)

Feature Flags

Global Flags

Troubleshooting

FAQ

Frequently Asked Questions

Is Codex CLI free?

What's the difference between Codex CLI and the Codex App?

Which model should I use with Codex?

What's the difference between codex exec and interactive codex?

How do I resume a previous session?

What does the sandbox actually do?

Can I use Codex CLI in my CI/CD pipeline?

How does Codex differ from Claude Code?

Finn Hillebrandt

Similar Articles

Claude Code Statistics 2026: Key Numbers, Data & Facts

OpenAI Codex Statistics 2026: Key Numbers, Data & Facts

How to Use Claude Code for Free (And What the Limits Are)

Claude Code Pricing: How Much Does Claude Code Really Cost?

Claude Code Alternatives 2026: 12 AI Coding Tools Compared

Claude Code: The Complete Beginner's Guide