GLM-ZAI-2API v1.0.0 - OpenAI Proxy for Z.AI

GLM-ZAI-2API v1.0.0

OpenAI-compatible proxy for Z.AI (GLM models) — single Go binary for all platforms.

🔥 Features

• 🚀 Single binary — server, captcha solver, and token store in one executable
• 🔓 OpenAI-compatible — /v1/chat/completions and /v1/models endpoints
• 🧩 Tool calling — full function calling support for Kiro and other IDEs
• 🛡️ In-process captcha — Aliyun Captcha V3 solver (no external dependencies)
• 📊 Embedded dashboard — live HTML status UI at http://localhost:5082
• 🌐 Cross-platform — Windows, macOS (Intel/ARM), Linux
• 📦 Standalone — pure-Go SQLite, zero CGO dependencies

💡 Works with Kiro out of the box! Perfect for AI-assisted development.

🚀 Supported Models

• glm-4.7 — works in guest mode (no token required)
• GLM-5-Turbo — requires ZAI_TOKEN
• GLM-5v-Turbo — requires ZAI_TOKEN (vision model)
• GLM-5.1 — requires ZAI_TOKEN
• glm-5.2 — requires ZAI_TOKEN (sometimes available)

📥 Downloads

Main API Server

Platform	Download	Size
Windows x64	glm-zai-2api-windows-amd64.exe	~10.4 MB
Linux x64	glm-zai-2api-linux-amd64	~10.2 MB
macOS Intel	glm-zai-2api-macos-amd64	~10.4 MB
macOS Apple Silicon	glm-zai-2api-macos-arm64	~9.8 MB

Token Collector

Platform	Download	Size
Windows x64	token-collector-windows-amd64.exe	~11.0 MB
Linux x64	token-collector-linux-amd64	~10.8 MB
macOS Intel	token-collector-macos-amd64	~11.0 MB
macOS Apple Silicon	token-collector-macos-arm64	~10.4 MB

🔧 Quick Start

1. Download

Download the appropriate binary for your platform from the assets below.

2. Collect Device Tokens

The captcha solver needs a pool of device tokens:

# Windows
.\token-collector-windows-amd64.exe --count 750 --out tokens.sqlite

# Linux/macOS
chmod +x ./token-collector-*
./token-collector-* --count 750 --out tokens.sqlite

3. Configure (Optional)

cp .env.example .env
# Edit .env to set ZAI_TOKEN for GLM-5 models

4. Run

# Windows
.\glm-zai-2api-windows-amd64.exe

# Linux/macOS
chmod +x ./glm-zai-2api-*
./glm-zai-2api-*

5. Access

• Dashboard: http://localhost:5082
• API: http://localhost:5082/v1/chat/completions

🔧 Kiro Configuration

Add as custom OpenAI endpoint in Kiro:

Base URL: http://localhost:5082/v1
API Key: d3vin
Model: glm-4.7 (guest) or GLM-5.1 (with token)

Full tool calling support works out of the box! 🎉

💡 Getting ZAI_TOKEN (for GLM-5 models)

1. Open https://chat.z.ai and log in
2. DevTools (F12) → Application → Local Storage → token
3. Copy the value (starts with eyJ...)
4. Add to .env: ZAI_TOKEN=eyJ...

📖 Documentation

• README: English | Русский
• Repository: https://github.com/D3-vin/GLM-ZAI-2API

📱 Contact

• Telegram: @D3_vin
• Author: @D3vin_dev
• GitHub: https://github.com/D3-vin/GLM-ZAI-2API

📝 What's New in v1.0.0

Initial release with:

• Complete OpenAI API compatibility
• Tool calling adapter for function calling
• In-process Aliyun Captcha V3 solver
• Cross-platform builds for 4 platforms
• Token collector utility
• Embedded web dashboard
• Session management with conversation history
• Multiple model support (GLM-4.7 to GLM-5.2)

---

License: Provided as-is for educational and interoperability purposes.