API申请
注册地址https://llm.leixin.me
step 1: 创建API密钥
step 2: 选择分组
step 3:复制密钥
codex 安装及使用
桌面端(macos)：
下载地址：https://chatgpt.com/zh-Hans-CN/download/
使用复制的密钥登录：
配置文件
打开访达：找到.codex mac下command + shift + . 显示隐藏文件
编辑./codex下的config.toml填入
toml
model_provider = "OpenAI"
model = "gpt-5.5"
review_model = "gpt-5.5"
model_reasoning_effort = "xhigh"
disable_response_storage = true
network_access = "enabled"
windows_wsl_setup_acknowledged = true
model_context_window = 1000000
model_auto_compact_token_limit = 900000
[model_providers.OpenAI]
name = "OpenAI"
base_url = "https://llm.leixin.me/v1"
wire_api = "responses"
requires_openai_auth = true重启codex即可
桌面端（windows):
下载地址：https://chatgpt.com/zh-Hans-CN/download/
使用复制的密钥登录
配置文件
显示隐藏文件
找到.codex目录
打开.codex中的config.toml填入
toml
model_provider = "OpenAI"
model = "gpt-5.5"
review_model = "gpt-5.5"
model_reasoning_effort = "xhigh"
disable_response_storage = true
network_access = "enabled"
windows_wsl_setup_acknowledged = true
model_context_window = 1000000
model_auto_compact_token_limit = 900000
[model_providers.OpenAI]
name = "OpenAI"
base_url = "https://llm.leixin.me/v1"
wire_api = "responses"
requires_openai_auth = true重启codex即可
vscode插件：
安装插件
填入配置：
配置文件
toml
model_provider = "OpenAI"
model = "gpt-5.5"
review_model = "gpt-5.5"
model_reasoning_effort = "xhigh"
disable_response_storage = true
network_access = "enabled"
windows_wsl_setup_acknowledged = true
model_context_window = 1000000
model_auto_compact_token_limit = 900000
[model_providers.OpenAI]
name = "OpenAI"
base_url = "https://llm.leixin.me/v1"
wire_api = "responses"
requires_openai_auth = true重启VScode即可
CLI安装(unbuntu):
安装npm
bash
sudo apt update
sudo apt install nodejs npm安装codex
bash
npm install -g @openai/codex运行codex并退出
bash
codex
ctrl+c填写配置文件
bash
vim ~/.codex/config.toml配置文件
bash
model_provider = "OpenAI"
model = "gpt-5.5"
review_model = "gpt-5.5"
model_reasoning_effort = "xhigh"
disable_response_storage = true
network_access = "enabled"
windows_wsl_setup_acknowledged = true
model_context_window = 1000000
model_auto_compact_token_limit = 900000
[model_providers.OpenAI]
name = "OpenAI"
base_url = "https://llm.leixin.me/v1"
wire_api = "responses"
requires_openai_auth = true