Grok-Api

A free Grok API wrapper that allows you to use Grok without API access or account authentication.

Overview

This project provides a Python-based API wrapper for Grok AI, enabling you to interact with Grok's conversational AI without requiring official API access or account credentials. It includes both a direct Python interface and a FastAPI server for easy integration into your applications.

Features

🔓 No Authentication Required - Access Grok without an account
🆓 Completely Free - No API keys or paid subscriptions needed
🚀 FastAPI Server - Ready-to-use REST API endpoint
🌐 Proxy Support - Full support for HTTP proxies
📡 Streaming Responses - Receive both complete responses and token-by-token streams
⚡ High Performance - Multi-worker support for concurrent requests

Installation

git clone https://github.com/realasfngl/Grok-Api.git
cd Grok-Api
pip install -r requirements.txt

Requirements

Python 3.10+
curl_cffi
fastapi
uvicorn
coincurve
beautifulsoup4
pydantic
colorama

Usage

Models:

Model	Mode	Description
`grok-3-auto`	auto	Automatic mode
`grok-3-fast`	fast	Fast processing mode
`grok-4`	expert	Expert mode
`grok-4-mini-thinking-tahoe`	grok-4-mini-thinking	Mini thinking mode

Manual Usage (Python)

New conversation:

from core import Grok

response = Grok("grok-3-fast").start_convo("Hello, how are you today?")
print(response)

proxy = "http://username:password@ip:port"
response = Grok("grok-3-fast", proxy).start_convo("Tell me a joke")
print(response)

Continue conversation:

from core import Grok

response = Grok().start_convo("Hello, how are you today?")
print(response)

response2 = Grok().start_convo("That's nice! Glad to hear!", extra_data=response["extra_data"])
print(response2)

Example Output:

{
    "response": "Yo, I'm just chilling in the digital realm...",
    "stream_response": ["Yo", ",", " I'm", " just", " chilling", "..."],
    "images": None,
    "extra_data": {"..."}
}

API Server

Starting the Server

Simple start:

python api_server.py

Production start with custom configuration:

uvicorn api_server:app --host 0.0.0.0 --port 6969 --workers 50

Making API Requests

New conversation:

import requests

response = requests.post(
    "http://localhost:6969/ask",
    json={
        "proxy": "http://user:pass@ip:port",
        "message": "Hello, Grok!",
        "model": "grok-3-fast",
        "extra_data": None
    }
)
print(response.json())

Continue conversation:

import requests

response1 = requests.post(
    "http://localhost:6969/ask",
    json={
        "proxy": "http://user:pass@ip:port",
        "message": "Hello!",
        "model": "grok-3-fast",
        "extra_data": None
    }
)
data1 = response1.json()
print(data1)

response2 = requests.post(
    "http://localhost:6969/ask",
    json={
        "proxy": "http://user:pass@ip:port",
        "message": "Tell me more",
        "model": "grok-3-fast",
        "extra_data": data1["extra_data"]
    }
)
print(response2.json())

API Response Format

{
  "status": "success",
  "response": "Complete response message from Grok",
  "stream_response": ["Token", "by", "token", "response", "array"],
  "images": null,
  "extra_data": {"..."}
}

Configuration

Proxy Format

The wrapper accepts proxies in the following formats:

http://ip:port
http://username:password@ip:port
ip:port (automatically prefixed with http://)

API Server Settings

Modify api_server.py to change:

Host: Default 0.0.0.0 (all interfaces)
Port: Default 6969
Workers: Default 50 (adjust based on your server capacity)

Troubleshooting

Common Issues:

IP Flag - {"error":{"code":7,"message":"Request rejected by anti-bot rules.","details":[]}} - This indicates your IP or proxy has been flagged. Try using a different proxy or IP address.

Support

If you find this project helpful, consider starring the repository!

Note: This project may break if Grok updates their web interface. Please report any issues if the wrapper stops working.

Contact: This project is for educational purposes only. If Grok has an Issue with this Project please contact me via my email nuhuh3116@gmail.com.