GlideflowAI
Open navigation
ModelsApps & AgentsPricingSetupDocsBlogLogin
Start free
← Apps & Agents
Verified setup

Use Codex CLI with GlideflowAI.

A configurable terminal agent for repository-sized coding work. Connect it through GlideflowAI's OpenAI-compatible API with a key you control.

Get a key

What this connection is for

Codex CLI is a terminal coding agent designed to work beside your local development tools. It can inspect a repository, propose a plan, edit files, and run verification commands while leaving the work visible in the shell and working tree. That makes it a practical option for developers who want agent assistance integrated with familiar review, test, and version-control habits.

The CLI separates the selected model from the configured provider. This is important when testing a compatible gateway: the provider block names the base URL and environment variable, while the top-level model and model_provider choose where a session is sent. A small configuration mistake can otherwise direct traffic to a default provider instead of the intended endpoint.

GlideflowAI's OpenAI-compatible endpoint is usable through a named model provider. Keep your key in an environment variable, not in the TOML file, and begin in a disposable or clean repository state. This is a user-managed connection. Codex CLI and GlideflowAI remain independent products.

Why pair it with GlideflowAI

One named provider and one key can serve Codex CLI alongside other OpenAI-compatible agents.

GlideflowAI keeps its catalog and pay-as-you-go USD pricing public, which is helpful when you decide what model to run for a repository task.

Setup

Use a new API key for the tool, select a model from the catalog, and begin with a small task. Do not put a key in a repository, screenshot, or shared configuration file.

  1. Add a provider block

    Open ~/.codex/config.toml and add a GlideflowAI model provider. The top-level model_provider value is required so Codex CLI selects this provider instead of its default.

    model = "glm-5.2"
model_provider = "glideflowai"

[model_providers.glideflowai]
name = "GlideflowAI"
base_url = "https://api.glideflowai.com/v1"
env_key = "GLIDEFLOW_API_KEY"

  2. Export your key

    Set the variable named by env_key in the shell that launches Codex CLI. Do not paste a live key into config.toml or a dotfile tracked by git.

    export GLIDEFLOW_API_KEY="sk-your-key"

  3. Launch and verify routing

    Run a modest task first, then check that the configured model provider is selected. Use the exact model ID shown in the catalog if you change models.

    codex

Need the endpoint basics? Read authentication, browse the model catalog, or compare pricing. For tool-specific troubleshooting, see the existing guide.

Frequently asked questions

Why is model_provider necessary?

Without it, the CLI can use its default provider rather than the GlideflowAI block you added.

Should I set wire_api?

No. The existing GlideflowAI guide documents the default transport; do not add a wire_api override for this setup.

Where is the full troubleshooting guide?

Open the Codex CLI tool guide in GlideflowAI Docs for a request-level endpoint test and common error notes.

Ready to connect Codex CLI?

Create a key, use the setup above, and verify the connection on a small task before you scale it up.

Get a key

Explore related agents

Claude CodeAiderOpenCode