> ## Documentation Index
> Fetch the complete documentation index at: https://docs.prisminference.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Codex

> Configure Codex to use Prism through the Responses API.

## Copy for your agent

```text theme={"theme":{"light":"github-light","dark":"github-dark"}}
Follow https://docs.prisminference.com/guides/codex to configure Prism in my
user-level Codex settings. Preserve unrelated settings and never repeat my API
key to me or include it in your response.
```

If no key exists, follow [Agent signup](/guides/agent-signup) first.

## Configure

Set `PRISM_API_KEY` in the environment that launches Codex. Merge this into
`~/.codex/config.toml`:

```toml theme={"theme":{"light":"github-light","dark":"github-dark"}}
model = "deepseek-v4.1-flash"
model_provider = "prism"
web_search = "disabled"

[model_providers.prism]
name = "Prism"
base_url = "https://api.prisminference.com/v1"
env_key = "PRISM_API_KEY"
wire_api = "responses"
```

Keep the provider at user scope; Codex ignores custom providers in project
configuration. Do not put the key in `config.toml`.

## Verify

```bash theme={"theme":{"light":"github-light","dark":"github-dark"}}
codex doctor --json
```

Codex sends the bare model ID `deepseek-v4.1-flash` to Prism's stateless
`POST /v1/responses` endpoint. Web search remains disabled because Prism does
not provide OpenAI-hosted tools.