DEV Community

TengLongAI2026
TengLongAI2026

Posted on

How I Gave Codex CLI Superpowers on Windows (with DeepSeek, No OpenAI Required)

#ai #opensource #tutorial #productivity

TL;DR: I installed Codex CLI on Windows, connected it to DeepSeek via a local shim server, and now I have a fully functional AI coding agent that doesn't need OpenAI at all. Here's exactly how I did it.

Why This Matters

I live in China. OpenAI's API is blocked here. For months, every time I tried to use an AI coding agent, I hit the same wall:

  • Codex CLI connects to OpenAI's WebSocket -> blocked
  • Claude Code needs an API key -> complicated
  • Cursor is paid -> expensive
  • GitHub Copilot -> limited model choice

But today, my boss shared a Toutiao article about codex-shim - a local Python server that intercepts Codex's API requests and routes them to any model you want.

I tried it. It worked. Here's the play-by-play.

The Architecture

Three components:

  1. Codex CLI - OpenAI's terminal coding agent (87K stars on GitHub)
  2. codex-shim - A local Python/aiohttp server that translates API calls
  3. DeepSeek Chat - The model doing the actual reasoning

Step 1: Install Codex CLI 

npm i -g @openai/codex
Enter fullscreen mode Exit fullscreen mode

5 seconds on Windows. Check: codex --version -> codex-cli 0.135.0

Step 2: Install codex-shim 

pip install ./codex-shim-0xSero/
Enter fullscreen mode Exit fullscreen mode

Requires Python 3.11+ and aiohttp. Works on Windows in PowerShell/cmd.

Step 3: Configure the Model

Create ~/.codex-shim/models.json with your DeepSeek API key and base URL.

Step 4: Start the Shim 

codex-shim --port 8765 start
curl http://127.0.0.1:8765/v1/models
Enter fullscreen mode Exit fullscreen mode

Step 5: Run Codex 

codex -p shim -s danger-full-access
Enter fullscreen mode Exit fullscreen mode

First Run

First test: I asked Codex to write a Fibonacci script in Python. It listed the directory, wrote the file, executed it, and verified output. 7,000+ tokens all routed through DeepSeek. No OpenAI needed.

Lessons Learned

  1. Don't accept vendor lock-in - there's always a shim
  2. Local-first is powerful - the shim runs on localhost
  3. Windows is not an afterthought - everything works
  4. The Chinese AI ecosystem is real - DeepSeek, Qwen, GLM are competitive

Resources

  • Codex CLI: github.com/openai/codex (87K stars)
  • codex-shim: github.com/0xSero/codex-shim (733 stars)
  • DeepSeek: platform.deepseek.com

From a computer in Shenzhen, running AI agents that don't care where the model lives. The future is multi-model.

Top comments (0)