Skip to content

Latest commit

 

History

History
328 lines (260 loc) · 9.48 KB

client.md

File metadata and controls

328 lines (260 loc) · 9.48 KB

G4F Client API Guide

Table of Contents

Introduction

Welcome to the G4F Client API, a cutting-edge tool for seamlessly integrating advanced AI capabilities into your Python applications. This guide is designed to facilitate your transition from using the OpenAI client to the G4F Client, offering enhanced features while maintaining compatibility with the existing OpenAI API.

Getting Started

Switching to G4F Client

To begin using the G4F Client, simply update your import statement in your Python code:

Old Import:

from openai import OpenAI

New Import:

from g4f.client import Client as OpenAI

The G4F Client preserves the same familiar API interface as OpenAI, ensuring a smooth transition process.

Initializing the Client

To utilize the G4F Client, create a new instance. Below is an example showcasing custom providers:

from g4f.client import Client
from g4f.Provider import BingCreateImages, OpenaiChat, Gemini

client = Client(
    provider=OpenaiChat,
    image_provider=Gemini,
    # Add any other necessary parameters
)

Creating Chat Completions

Here’s an improved example of creating chat completions:

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[
        {
            "role": "user",
            "content": "Say this is a test"
        }
    ]
    # Add any other necessary parameters
)

This example:

  • Asks a specific question Say this is a test
  • Configures various parameters like temperature and max_tokens for more control over the output
  • Disables streaming for a complete response

You can adjust these parameters based on your specific needs.

Configuration

You can set an api_key for your provider in the client and define a proxy for all outgoing requests:

from g4f.client import Client

client = Client(
    api_key="your_api_key_here",
    proxies="http://user:pass@host",
    # Add any other necessary parameters
)

Explanation of Parameters

When using the G4F to create chat completions or perform related tasks, you can configure the following parameters:

  • model:
    Specifies the AI model to be used for the task. Examples include "gpt-4o" for GPT-4 Optimized or "gpt-4o-mini" for a lightweight version. The choice of model determines the quality and speed of the response. Always ensure the selected model is supported by the provider.

  • messages:
    A list of dictionaries representing the conversation context. Each dictionary contains two keys: - role: Defines the role of the message sender, such as "user" (input from the user) or "system" (instructions to the AI).
    - content: The actual text of the message.
    Example:

    [
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What day is it today?"}
    ]
  • web_search:
    (Optional) A Boolean flag indicating whether to enable internet-based search capabilities for the task. If True, the system performs a web search using the DuckDuckGo search engine to retrieve up-to-date information. This is particularly useful for obtaining real-time or specific details not contained within the model's training.

  • provider:
    Specifies the backend provider for the API. Examples include g4f.Provider.Blackbox or g4f.Provider.OpenaiChat. Each provider may support a different subset of models and features, so select one that matches your requirements.

Usage Examples

Text Completions

Generate text completions using the ChatCompletions endpoint:

from g4f.client import Client

client = Client()

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[
        {
            "role": "user",
            "content": "Say this is a test"
        }
    ],
    web_search = False
    # Add any other necessary parameters
)

print(response.choices[0].message.content)

Streaming Completions

Process responses incrementally as they are generated:

from g4f.client import Client

client = Client()

stream = client.chat.completions.create(
    model="gpt-4",
    messages=[
        {
            "role": "user",
            "content": "Say this is a test"
        }
    ],
    stream=True,
)

for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content or "", end="")

Image Generation

The response_format parameter is optional and can have the following values:

  • If not specified (default): The image will be saved locally, and a local path will be returned (e.g., "/images/1733331238_cf9d6aa9-f606-4fea-ba4b-f06576cba309.jpg").
  • "url": Returns a URL to the generated image.
  • "b64_json": Returns the image as a base64-encoded JSON string.

Generate images using a specified prompt:

from g4f.client import Client

client = Client()

response = client.images.generate(
    model="flux",
    prompt="a white siamese cat",
    response_format="url"
    # Add any other necessary parameters
)

image_url = response.data[0].url

print(f"Generated image URL: {image_url}")

Base64 Response Format

from g4f.client import Client

client = Client()

response = client.images.generate(
    model="flux",
    prompt="a white siamese cat",
    response_format="b64_json"
    # Add any other necessary parameters
)

base64_text = response.data[0].b64_json
print(base64_text)

Creating Image Variations

Create variations of an existing image:

from g4f.client import Client
from g4f.Provider import OpenaiChat

client = Client(
    image_provider=OpenaiChat
)

response = client.images.create_variation(
    image=open("docs/images/cat.jpg", "rb"),
    model="dall-e-3",
    # Add any other necessary parameters
)

image_url = response.data[0].url

print(f"Generated image URL: {image_url}")

Advanced Usage

Using a List of Providers with RetryProvider

from g4f.client import Client
from g4f.Provider import RetryProvider, Phind, FreeChatgpt, Liaobots
import g4f.debug

g4f.debug.logging = True
g4f.debug.version_check = False

client = Client(
    provider=RetryProvider([Phind, FreeChatgpt, Liaobots], shuffle=False)
)

response = client.chat.completions.create(
    model="",
    messages=[
        {
            "role": "user",
            "content": "Hello"
        }
    ]
)

print(response.choices[0].message.content)

Using a Vision Model

Analyze an image and generate a description:

import g4f
import requests

from g4f.client import Client
from g4f.Provider.GeminiPro import GeminiPro

# Initialize the GPT client with the desired provider and api key
client = Client(
    api_key="your_api_key_here",
    provider=GeminiPro
)

image = requests.get("https://raw.githubusercontent.com/xtekky/gpt4free/refs/heads/main/docs/cat.jpeg", stream=True).raw
# Or: image = open("docs/images/cat.jpeg", "rb")

response = client.chat.completions.create(
    model=g4f.models.default,
    messages=[
        {
            "role": "user",
            "content": "What are on this image?"
        }
    ],
    image=image
    # Add any other necessary parameters
)

print(response.choices[0].message.content)

Command-line Chat Program

Here's an example of a simple command-line chat program using the G4F Client:

import g4f
from g4f.client import Client

# Initialize the GPT client with the desired provider
client = Client()

# Initialize an empty conversation history
messages = []

while True:
    # Get user input
    user_input = input("You: ")

    # Check if the user wants to exit the chat
    if user_input.lower() == "exit":
        print("Exiting chat...")
        break  # Exit the loop to end the conversation

    # Update the conversation history with the user's message
    messages.append({"role": "user", "content": user_input})

    try:
        # Get GPT's response
        response = client.chat.completions.create(
            messages=messages,
            model=g4f.models.default,
        )

        # Extract the GPT response and print it
        gpt_response = response.choices[0].message.content
        print(f"Bot: {gpt_response}")

        # Update the conversation history with GPT's response
        messages.append({"role": "assistant", "content": gpt_response})

    except Exception as e:
        print(f"An error occurred: {e}")

This guide provides a comprehensive overview of the G4F Client API, demonstrating its versatility in handling various AI tasks, from text generation to image analysis and creation. By leveraging these features, you can build powerful and responsive applications that harness the capabilities of advanced AI models.


Return to Home