Augment2Api

Using this item will quickly result in your account being banned, so it is no longer recommended.

English | 中文

A middleware service for connecting to Augment API, providing OpenAI-compatible interfaces with support for Claude3.7、Claude4 model calls.

📋 Table of Contents

• Features
• Important Notice
• Supported Models
• Environment Variables
• Quick Start
• API Usage
• Admin Interface
• Batch Add Tokens
• Feedback
• Star History

✨ Features

• 🔄 OpenAI-compatible API interface
• 🤖 Support for Claude-Sonnet-3.7 model(In Chat Mode),Cluade-4-Sonnet(In Agent Mode)
• 📡 Support for streaming/non-streaming output
• 🎛️ Simple multi-token management interface
• 🗄️ Redis-based token storage
• 🔍 Batch token and tenant URL detection and updates
• 📥 API for batch token addition

⚠️ Important Notice

Risk Warning: Using this project may result in your account being flagged, restricted, or banned. Use at your own risk!

• The system determines the usage mode based on the input model name. In AGENT mode, all tool calls are blocked, using the model's native capabilities to respond, otherwise conversations may be interrupted by tool calls
• Default concurrency control is applied, limiting each token to a maximum of 1 request every 3 seconds, with default Block Token cooldown rules
• Augment's Agent mode is very powerful. We recommend using the official plugin in your editor for an experience comparable to Cursor

🤖 Supported Models

Model names ending with -chat use CHAT mode for responses

Model names ending with -agent use AGENT mode for responses  

Other model names default to CHAT mode

🔧 Environment Variables

Variable	Description	Required	Example
REDIS_CONN_STRING	Redis connection string	✅ Yes	redis://default:password@localhost:6379
ACCESS_PWD	Admin panel access password	✅ Yes	your-access-password
AUTH_TOKEN	API access authentication token	❌ No	your-auth-token
ROUTE_PREFIX	API request prefix	❌ No	your_api_prefix
CODING_MODE	Debug mode switch	❌ No	false
CODING_TOKEN	Debug token	❌ No	empty
TENANT_URL	Debug tenant URL	❌ No	empty
PROXY_URL	HTTP proxy address	❌ No	http://127.0.0.1:7890
REMOVE_FREE	Remove free accounts switch	❌ No	false

Tip: If the page fails to get tokens, you can set CODING_MODE=true and configure CODING_TOKEN and TENANT_URL to use a specific token and tenant URL (limited to single token usage).

REMOVE_FREE Environment Variable

Since Augment doesn't automatically switch to Free plan after trial ends, causing conversations to respond with plan switching prompts, if you need to automatically remove these accounts, you can set REMOVE_FREE=true. The system will automatically identify and disable free accounts during batch detection.

If the response contains the following content:

• "Your subscription for account"
• "is inactive"
• "update your plan here to continue using Augment"

The token will be automatically marked as unavailable and will not participate in subsequent API call allocations.

This helps improve API success rates and stability, avoiding request failures due to free account limitations.

🚀 Quick Start

Option 1: Using Docker

docker run -d \
  --name augment2api \
  -p 27080:27080 \
  -e REDIS_CONN_STRING="redis://default:yourpassword@your-redis-host:6379" \
  -e ACCESS_PWD="your-access-password" \
  -e AUTH_TOKEN="your-auth-token" \
  --restart always \
  linqiu1199/augment2api

Option 2: Using Docker Compose

1. Clone the project

git clone https://github.com/linqiu1199/augment2api.git
cd augment2api

2. Configure environment variables

Create a .env file:

# Set Redis password (required)
REDIS_PASSWORD=your-redis-password

# Set admin panel access password (required)
ACCESS_PWD=your-access-password

# Set api authentication token (optional)
AUTH_TOKEN=your-auth-token

3. Start services

docker-compose up -d

This will start both Redis and Augment2Api services and automatically handle network connections between them.

Getting Tokens

1. Visit http://ip:27080/ to access the admin login page, enter your access password to enter the admin panel
2. Click the Add TOKEN menu

3. Follow these steps to get tokens:
   • Click to get authorization link
   • Copy the authorization link and open it in browser
   • Login with email (domain email also works)
   • Copy the augment code to the authorization response input box, click get token
   • Token should appear normally in the TOKEN list

4. Start conversation testing

Tips:

• If conversation returns 503 error, run batch detection once to update tenant URLs before testing conversations (tenant URL error)
• If conversation returns 401 error, run batch detection once to disable invalid tokens before testing conversations (account banned)

📖 API Usage

Get Model List

curl -X GET http://localhost:27080/v1/models

Chat Interface

curl -X POST http://localhost:27080/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
  "model": "claude-3.7",
  "messages": [
    {"role": "user", "content": "Hello, please introduce yourself"}
  ]
}'

🎛️ Admin Interface

Visit http://localhost:27080/ to open the admin login page. After logging in, you can interactively get and manage tokens.

📥 Batch Add Tokens

Without AUTH_TOKEN set

curl -X POST http://localhost:27080/api/add/tokens \
-H "Content-Type: application/json" \
-d '[
  {
    "token": "token1",
    "tenantUrl": "https://tenant1.com"
  },
  {
    "token": "token2",
    "tenantUrl": "https://tenant2.com"
  }
]'

With AUTH_TOKEN set

curl -X POST http://localhost:27080/api/add/tokens \
-H "Content-Type: application/json" \
-H "Authorization: Bearer your-auth-token" \
-d '[
  {
    "token": "token1",
    "tenantUrl": "https://tenant1.com"
  },
  {
    "token": "token2",
    "tenantUrl": "https://tenant2.com"
  }
]'    

💬 Feedback

🐞 Telegram Group

📈 Star History

---

If this project helps you, please give us a ⭐ Star!