Codex

Codex is built for code generation, repository analysis, automated edits, and command-line workflows. XNova's OpenAI groups support the Responses format (/v1/responses) that Codex uses.

Using the Codex desktop app and prefer a GUI for providers? See Codex++.

config.toml

Edit ~/.codex/config.toml:

model = "gpt-6.1-sol"
model_provider = "xnova"

[model_providers.xnova]
name = "XNova"
base_url = "https://api.xnova.online/v1"
env_key = "XNOVA_API_KEY"
wire_api = "responses"

Then set the key and start Codex:

export XNOVA_API_KEY="sk-your-xnova-key"
codex

Common models

gpt-6.1-sol
gpt-6-astra
gpt-6-sol
gpt-5.6-sol

Verify

curl https://api.xnova.online/v1/models \
  -H "Authorization: Bearer sk-your-xnova-key"

Troubleshooting

• Model not found: sync the model list or enter the exact model name.
• 401: make sure the environment variable name matches env_key; recreate the key if needed.
• Slow requests: try https://cf-api.xnova.online/v1 and check client timeouts.