# Codex CLI (/docs/code-cli/codex)

[Codex CLI](https://github.com/openai/codex) runs on OhMyGPT with one provider entry in its config file. Your OhMyGPT key then gives Codex the GPT models and the rest of the catalog, switchable per run with `codex -m`. You can test the whole setup for free with `ohmygpt/mock`.

## Configure [#configure]

Add the provider to `~/.codex/config.toml`:

```toml title="~/.codex/config.toml"
model = "openai/gpt-5.5"
model_provider = "ohmygpt"

[model_providers.ohmygpt]
name = "OhMyGPT"
base_url = "https://api.ohmygpt.com/v1"
env_key = "OHMYGPT_API_KEY"
wire_api = "responses"
```

Codex reads your key from the variable named in `env_key`. Export it in the shell that runs Codex:

```bash
export OHMYGPT_API_KEY="sk-..."
codex
```

- `base_url` ends in `/v1`. Codex appends `/responses`.
- `wire_api = "responses"` selects the Responses API, which Codex uses for every provider.
- `model` is any model ID from [Choose a model](#choose-a-model). To use another model for one run, start Codex with `codex -m <model id>`.

## Check the connection [#check-the-connection]

Send a Responses request, the kind Codex sends, to the free test model `ohmygpt/mock`:

```bash
curl https://api.ohmygpt.com/v1/responses \
  -H "Authorization: Bearer $OHMYGPT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "ohmygpt/mock", "input": "ping"}'
```

```json
{
  "id": "resp_MRazQMjAopUMBIjoOfs2bSUL",
  "object": "response",
  "status": "completed",
  "model": "mock",
  "output": [
    {
      "type": "message",
      "id": "msg_0",
      "role": "assistant",
      "status": "completed",
      "content": [
        {
          "type": "output_text",
          "text": "This is a mock reply from OhMyGPT. Your integration works; switch to a real model when ready."
        }
      ]
    }
  ],
  "usage": { "input_tokens": 15, "output_tokens": 17, "total_tokens": 32 }
}
```

To run Codex itself at no cost, start it with `codex -m ohmygpt/mock`. The mock calls the first tool Codex declares, gets the result back and replies with text, so you see one full turn of the agent loop. The mock is available once your account has been topped up.

## Choose a model [#choose-a-model]

List the models you can use with Codex:

```bash
curl -s https://api.ohmygpt.com/v1/models \
  | jq -r '.data[] | select(.available and any(.protocols[]?; .id == "openai-responses")) | .id'
```

Prices for each are on the [Models](https://www.ohmygpt.com/models) page.

Codex is tuned for OpenAI's GPT models, and `openai/gpt-*` IDs give you the full experience. Other models in the list, such as `zhipu/glm-5.3`, run Codex too, including its tool calls and reasoning settings.

## Troubleshooting [#troubleshooting]

| Symptom | Cause | Fix |
|---|---|---|
| ``Missing environment variable: `OHMYGPT_API_KEY`.`` | The variable is not exported in the shell that runs Codex | `export OHMYGPT_API_KEY=...`, then start Codex again |
| `` `wire_api = "chat"` is no longer supported. `` | The provider entry uses an older Codex setting | Set `wire_api = "responses"` |
| Every request returns `404` | `base_url` is missing `/v1`, so requests go to `/responses` | Set `base_url = "https://api.ohmygpt.com/v1"` |
| `401` with `Invalid API key` | The key is mistyped or was deleted | Copy a key from [API keys](https://www.ohmygpt.com/keys) |
| `404` saying the model does not exist | The ID is misspelled | Copy the ID from `GET /v1/models` |
| `402` | The balance or the key's spending limit is used up | Top up at [Billing](https://www.ohmygpt.com/billing) |

Every response carries an `X-Request-ID` header, and [Usage](https://www.ohmygpt.com/usage) shows what each request cost.