Claude Code API Configuration

Claude's API isn't very friendly to use in China. Apart from Claude's official models, can I use other AI models?

This chapter will teach you in detail how to configure APIs for multiple mainstream domestic AI models, making Claude Code support:

Vendor/Brand Introduction API Application Portal (Click to Reach)
DeepSeek (Domestic High Cost-Performance) Official Models:
deepseek-v4-flash
deepseek-v4-pro
deepseek-chat (will be deprecated on 2026/07/24)
deepseek-reasoner (will be deprecated on 2026/07/24)		 https://platform.deepseek.com/api_keys
Alibaba Bailian (Qwen) Alibaba Cloud's unified LLM gateway, supports the latest versions of Qwen, GLM, Kimi, MiniMax, etc. https://bailian.console.aliyun.com
GLM (Zhipu Qingyan) Tsinghua-affiliated ChatGLM series, supports GLM-4, GLM-3-Turbo, etc. https://open.bigmodel.cn
MiniMax Domestic multimodal, supports mixed calls of text, speech, and images https://platform.minimaxi.com

After entering the corresponding console, register/log in → complete real-name verification → create an API Key and you can start calling.

API Management Tool

With so many platforms, configuration becomes troublesome. We can use the third-party tool CC Switch to easily manage the API configurations of these popular tools:https://github.com/farion1231/cc-switch/, supports Windows / macOS / Linux.

CC Switch is a comprehensive auxiliary tool for Claude Code / Codex / Gemini CLI.

CC Switch can help us easily manage the API configurations of these popular tools, just like having a smart organization assistant for your development toolbox, where all tool configurations can be managed in an orderly manner.

Installation package download addresses for each platform:https://github.com/farion1231/cc-switch/releases。

For specific operation settings, refer to the article:https://www.example.com/ai-agent/cc-switch.html

If you don't mind the hassle, you can refer to the following content and configure it yourself.

Integrating DeepSeek with Claude Code

Information Needed:

Configure in Environment Variables

Linux / Mac users run the following commands to configure the DeepSeek Anthropic API environment variables, where the API Key can be obtained atDeepSeek PlatformGet it at:

export ANTHROPIC_BASE_URL=https://api.deepseek.com/anthropic
export ANTHROPIC_AUTH_TOKEN=<你的 DeepSeek API Key>
export ANTHROPIC_MODEL=deepseek-v4-pro[1m]
export ANTHROPIC_DEFAULT_OPUS_MODEL=deepseek-v4-pro[1m]
export ANTHROPIC_DEFAULT_SONNET_MODEL=deepseek-v4-pro[1m]
export ANTHROPIC_DEFAULT_HAIKU_MODEL=deepseek-v4-flash
export CLAUDE_CODE_SUBAGENT_MODEL=deepseek-v4-flash
export CLAUDE_CODE_EFFORT_LEVEL=max

Windows users run:

$env:ANTHROPIC_BASE_URL="https://api.deepseek.com/anthropic"
$env:ANTHROPIC_AUTH_TOKEN="<你的 DeepSeek API Key>"
$env:ANTHROPIC_MODEL="deepseek-v4-pro[1m]"
$env:ANTHROPIC_DEFAULT_OPUS_MODEL="deepseek-v4-pro[1m]"
$env:ANTHROPIC_DEFAULT_SONNET_MODEL="deepseek-v4-pro[1m]"
$env:ANTHROPIC_DEFAULT_HAIKU_MODEL="deepseek-v4-flash"
$env:CLAUDE_CODE_SUBAGENT_MODEL="deepseek-v4-flash"
$env:CLAUDE_CODE_EFFORT_LEVEL="max"

Note: Add [1m] to the model configuration, and remember to set thinking intensity to max, so that it can run at full capacity.

Parameter Description:

Simplified to the first three columns (Parameter Name / Example Value / Description) as follows:

Parameter Name Example Value Description
ANTHROPIC_BASE_URL https://api.deepseek.com/anthropic Specifies the API gateway address, forwarding Anthropic SDK requests to DeepSeek
ANTHROPIC_AUTH_TOKEN <你的 DeepSeek API Key> API authentication Token, used to access DeepSeek services
ANTHROPIC_MODEL deepseek-v4-pro The default main model used
ANTHROPIC_DEFAULT_OPUS_MODEL deepseek-v4-pro Maps to the Claude Opus model
ANTHROPIC_DEFAULT_SONNET_MODEL deepseek-v4-pro Maps to the Claude Sonnet model
ANTHROPIC_DEFAULT_HAIKU_MODEL deepseek-v4-flash Maps to the Claude Haiku model (lightweight and fast)
CLAUDE_CODE_SUBAGENT_MODEL deepseek-v4-flash The model used by sub-agents (usually for quick tasks)
CLAUDE_CODE_EFFORT_LEVEL max Controls reasoning intensity and thinking depth

Then enter the project directory, run the claude command, and you can start using it.

cd my-project
claude

After starting Claude Code, specify DeepSeek:

claude --model deepseek-v4-pro

Or switch in interactive mode:

> /model deepseek-v4-pro

Enter /model to view supported models:

Reference documentation:https://api-docs.deepseek.com/zh-cn/guides/coding_agents

Integrating Alibaba Bailian with Claude Code

Alibaba Cloud Bailian's Tongyi Qianwen series models support the Anthropic API-compatible interface. By modifying the following parameters, you can call Tongyi Qianwen series models in Claude Code.

macOS/Linux：

Create and open the configuration file~/.claude/settings.json。

~Represents the user's home directory. If.claudethe directory does not exist, you need to create it first. You can run the following command in the terminal:mkdir -p ~/.claude to create it.

Edit the configuration file:

vim ~/.claude/settings.json

Replace YOUR_API_KEY with your dedicated API Key.

{
    "env": {
        "ANTHROPIC_AUTH_TOKEN": "YOUR_API_KEY",
        "ANTHROPIC_BASE_URL": "https://dashscope.aliyuncs.com/apps/anthropic",
        "ANTHROPIC_MODEL": "qwen3-coder-plus"
    }
}

Open a new terminal again to make the environment variable configuration take effect.

Windows：

Run the following commands in CMD to set environment variables:

setx ANTHROPIC_AUTH_TOKEN "YOUR_API_KEY"
setx ANTHROPIC_BASE_URL "https://dashscope.aliyuncs.com/apps/anthropic"
setx ANTHROPIC_MODEL "qwen3-coder-plus"

Replace YOUR_API_KEY with the plan-specific API Key.

Open a new CMD window and run the following command to check whether the environment variable has taken effect.

echo %ANTHROPIC_AUTH_TOKEN%
echo %ANTHROPIC_BASE_URL%
echo %ANTHROPIC_MODEL%

Run the following commands in PowerShell to set environment variables:

# 用套餐专属 API Key 替换 YOUR_API_KEY
[Environment]::SetEnvironmentVariable("ANTHROPIC_AUTH_TOKEN", "YOUR_API_KEY", [EnvironmentVariableTarget]::User)
[Environment]::SetEnvironmentVariable("ANTHROPIC_BASE_URL", "https://dashscope.aliyuncs.com/apps/anthropic", [EnvironmentVariableTarget]::User)
[Environment]::SetEnvironmentVariable("ANTHROPIC_MODEL", "qwen3-coder-plus", [EnvironmentVariableTarget]::User)

Open a new PowerShell window and run the following command to check whether the environment variable has taken effect.

echo $env:ANTHROPIC_AUTH_TOKEN
echo $env:ANTHROPIC_BASE_URL
echo $env:ANTHROPIC_MODEL

Integrating Qwen Series Models in Claude Code

During the conversation, run/model <model name>command to switch the model.

/model qwen3-coder-plus

You can also create a file in the project root directory,.claude/settings.jsonand write the model configuration information into it for permanent configuration.

{
  "env": {
    "ANTHROPIC_MODEL": "qwen3-coder-plus",
    "ANTHROPIC_SMALL_FAST_MODEL": "qwen-flash"
  }
}

Start Claude and you can see the configuration information:

Integrating Zhipu Large Models with Claude Code

In this section, we use the ~/.claude/settings.json file to configure the large model. Before starting, you need to obtain an API key from the official platform:GLM Coding Plan。

Edit or add the Claude Code configuration file ~/.claude/settings.json, and add or modify the env field in it

# 注意替换里面的your_zhipu_api_key 为您上一步获取到的 API Key
{
    "env": {
        "ANTHROPIC_AUTH_TOKEN": "your_zhipu_api_key",
        "ANTHROPIC_BASE_URL": "https://open.bigmodel.cn/api/anthropic",
        "API_TIMEOUT_MS": "3000000",
        "CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC": 1
    }
}

Run claude to start Claude Code, enter/statusto confirm the model status:

If not, you can enter/configto switch the model.

Other extensions