Skip to main content
Codex CLI can talk to Firmware by setting a custom base URL. This lets you use models from multiple providers with one key.

​
Acknowledge

Setup notes are adapted from guidance by AlexGS74. Big thanks for sharing the original walkthrough.

​
Install

Install the Codex CLI. 
npm install -g @openai/codex
Confirm it’s on your PATH. 
codex --version

​
Configure

Create or edit ~/.codex/config.toml. Set a provider that points to Firmware. 
[model_providers.firmware]
name = "Firmware"
base_url = "https://app.firmware.ai/api/v1"
env_key = "FIRMWARE_API_KEY"

[profiles.firmware]
model_provider = "firmware"
model = "gpt-4o"
Export your API key in your shell. 
export FIRMWARE_API_KEY="fw_api_..."

​
Pick a model

Choose a model ID from the models list. You can set it in the profile or pass --model per run.

​
Run

Use the Firmware profile for any session. 
codex --profile firmware
Override the model when you need to. 
codex --profile firmware --model "claude-opus-4-5"

​
Troubleshoot

If you get a 401, double-check the FIRMWARE_API_KEY value. If you get a 404, make sure base_url ends with /v1.

​
Learn more

Models

Pick a supported model ID

Codex repo

View the CLI docs and config options