NextChat (ChatGPT-Next-Web)

NextChat is a widely-used open-source ChatGPT web client that supports custom OpenAI-compatible endpoints.

Official repository: https://github.com/ChatGPTNextWeb/NextChat Official docs: https://docs.nextchat.dev/quickstart/selfhost/env

UI configuration

1. Click the Settings icon in the lower left
2. Find the Custom Endpoint toggle in the settings panel and turn it on
3. Fill in the connection details:

Field	Value
OpenAI API Key	Your codex_... or sk-... key
API Proxy URL	https://codex.openai.com/v1

4. Save and start a conversation to verify.

Self-hosted deployment

For self-hosted instances, configure via environment variables:

OPENAI_API_KEY=codex_your_api_key
BASE_URL=https://codex.openai.com/v1

Restart the service after setting the variables.

Custom models

In the settings panel's Model dropdown, you can manually enter gpt-5-codex-mini or any other available model name.

For self-hosted deployments, you can also pre-set the model list via the CUSTOM_MODELS environment variable:

CUSTOM_MODELS=gpt-5-codex-mini

Notes

• The API Proxy URL takes the base URL including /v1: https://codex.openai.com/v1
• Available models can be found via GET /v1/models