CLI Integration
1
2
Configure Codex config file
Update your
$CODEX_HOME/config.toml file to include the Planck provider configuration:$CODEX_HOME is typically ~/.codex on Mac or Linux.config.toml
3
Set your Planck API key
Set the
PLANCK_API_KEY environment variable:4
Run Codex with Planck
Use Codex as normal. Your requests will automatically be logged to Planck:
5
SDK Integration
1
2
Install the Codex SDK
3
Configure the SDK with Planck
Initialize the Codex SDK with the AI Gateway base URL:
The Codex SDK doesn’t currently support specifying the wire API, so it will use the Responses API by default. This works with the AI Gateway with limited model and provider support. See the Responses API documentation for more details.
4
VS Code extension and remote environments
In the Codex extension, select Settings → Open config.toml. The default is~/.codex/config.toml (%USERPROFILE%\.codex\config.toml on Windows); when
CODEX_HOME is set, use $CODEX_HOME/config.toml. Merge provider settings into
the existing file, keeping top-level model settings before TOML tables.
Make PLANCK_API_KEY available to the process running Codex. Fully exit VS Code
and launch it from the configured terminal for local setup; launching from the
Dock or Start menu does not inherit those exports. For WSL, SSH, or containers,
configure the environment and user config of the host running the extension.
Do not place keys in TOML or commit them to a repository.
Use an approved model recognized by the installed client. Custom model aliases
require a matching administrator-supplied model catalog. Gateway route registration
does not establish streaming, tool-call, or continuation compatibility.
In Planck’s Connect your coding tools journey, copy the generated setup header
into http_headers under [model_providers.planck]. Send the displayed prompt
from a new Codex conversation and wait for the matching request in Planck. A
browser gateway test is separate from tool verification.
See official gateway setup
and compatibility requirements.