gpt4free/README.md
Heiner Lohaus e5b7f72b71 Move some modules, create providers dir
Set min version for duckduckgo
Make duckduckgo search async
Remove get_lastet_version
2024-02-22 00:16:58 +01:00

33 KiB
Raw Blame History

248433934-7886223b-c1d1-4260-82aa-da5741f303bb Written by @xtekky & maintained by @hlohaus

By using this repository or any code related to it, you agree to the legal notice. The author is not responsible for any copies, forks, re-uploads made by other users, or anything else related to GPT4Free. This is the author's only account and repository. To prevent impersonation or irresponsible actions, please comply with the GNU GPL license this Repository uses.

Note

Lastet version: PyPI version Docker version
Stats: Downloads Downloads

pip install -U g4f
docker pull hlohaus789/g4f

🆕 What's New

Feedback

You can always leave some feedback here: https://forms.gle/FeWV9RLEedfdkmFN6

To do

As per the survey, here is a list of improvements to come

  • golang implementation
  • Improve Documentation (on g4f.mintlify.app) & Do video tutorials
  • Improve the provider status list & updates
  • Tutorials on how to reverse sites to write your own wrapper (PoC only ofc)
  • Improve the Bing wrapper. (might write a new wrapper in golang as it is very fast)
  • Write a standard provider performance test to improve the stability
  • update the repository to include the new openai library syntax (ex: Openai() class)
  • Potential support and development of local models
  • improve compatibility and error handling

📚 Table of Contents

🛠️ Getting Started

Docker container

Quick start:
  1. Download and install Docker
  2. Pull latest image and run the container:
docker pull hlohaus789/g4f
docker run -p 8080:8080 -p 1337:1337 -p 7900:7900 --shm-size="2g" hlohaus789/g4f:latest
  1. Open the included client on: http://localhost:8080/chat/ or set the api base in your client to: http://localhost:1337/v1
  2. (Optional) If you need to log in to a provider, you can view the desktop from the container here: http://localhost:7900/?autoconnect=1&resize=scale&password=secret.

Use python

Prerequisites:
  1. Download and install Python (Version 3.10+ is recommended).
  2. Install Google Chrome for providers with webdriver
Install using pypi:
pip install -U g4f[all]

Or use partial requirements.

See: /docs/requirements

Install from source:

See: /docs/git

Install using Docker

See: /docs/docker

💡 Usage

Text Generation

with Python

from g4f.client import Client

client = Client()
response = client.chat.completions.create(
    model="gpt-3.5-turbo",
    messages=[{"role": "user", "content": "Say this is a test"}],
    ...
)
print(response.choices[0].message.content)

Image Generation

with Python

from g4f.client import Client

client = Client()
response = client.images.generate(
  model="gemini",
  prompt="a white siamese cat",
  ...
)
image_url = response.data[0].url

Result:

Image with cat

See also for Python:

Web UI

To start the web interface, type the following codes in python:

from g4f.gui import run_gui
run_gui()

or execute the following command:

python -m g4f.cli gui -port 8080 -debug

Interference API

You can use the Interference API to serve other OpenAI integrations with G4F.

See: /docs/interference

Configuration

Cookies / Access Token

For generating images with Bing and for the OpenAi Chat you need cookies or a token from your browser session. From Bing you need the "_U" cookie and from OpenAI you need the "access_token". You can pass the cookies / the access token in the create function or you use the set_cookies setter before you run G4F:

from g4f.cookies import set_cookies

set_cookies(".bing.com", {
  "_U": "cookie value"
})
set_cookies("chat.openai.com", {
  "access_token": "token value"
})
set_cookies(".google.com", {
  "__Secure-1PSID": "cookie value"
})

...

Alternatively, G4F reads the cookies with browser_cookie3 from your browser or it starts a browser instance with selenium webdriver for logging in.

Using Proxy

If you want to hide or change your IP address for the providers, you can set a proxy globally via an environment variable:

  • On macOS and Linux:
export G4F_PROXY="http://host:port"
  • On Windows:
set G4F_PROXY=http://host:port

🚀 Providers and Models

GPT-4

Website Provider GPT-3.5 GPT-4 Stream Status Auth
bing.com g4f.Provider.Bing ✔️ ✔️ Active
free.chatgpt.org.uk g4f.Provider.FreeChatgpt ✔️ ✔️ ✔️ Unknown
liaobots.site g4f.Provider.Liaobots ✔️ ✔️ ✔️ Active
chat.openai.com g4f.Provider.OpenaiChat ✔️ ✔️ ✔️ Active ✔️
raycast.com g4f.Provider.Raycast ✔️ ✔️ ✔️ Unknown ✔️
beta.theb.ai g4f.Provider.Theb ✔️ ✔️ ✔️ Unknown
chat.geekgpt.org g4f.Provider.GeekGpt ✔️ ✔️ ✔️ Inactive

GPT-3.5

Website Provider GPT-3.5 GPT-4 Stream Status Auth
chat3.aiyunos.top g4f.Provider.AItianhuSpace ✔️ ✔️ Unknown
aichatonline.org g4f.Provider.AiChatOnline ✔️ ✔️ Unknown
openchat.team g4f.Provider.Aura ✔️ ✔️ Active
chatbase.co g4f.Provider.ChatBase ✔️ ✔️ Active
chatforai.store g4f.Provider.ChatForAi ✔️ ✔️ Unknown
chatgpt.ai g4f.Provider.ChatgptAi ✔️ ✔️ Unknown
chat.chatgptdemo.net g4f.Provider.ChatgptDemo ✔️ ✔️ Unknown
chatgpt-free.cc g4f.Provider.ChatgptNext ✔️ ✔️ Unknown
chat.3211000.xyz g4f.Provider.Chatxyz ✔️ ✔️ Unknown
gptalk.net g4f.Provider.GPTalk ✔️ ✔️ Unknown
geminiprochat.com g4f.Provider.GeminiProChat ✔️ ✔️ Active
gpt6.ai g4f.Provider.Gpt6 ✔️ ✔️ Unknown
gptchatly.com g4f.Provider.GptChatly ✔️ Unknown
ai18.gptforlove.com g4f.Provider.GptForLove ✔️ ✔️ Unknown
gptgo.ai g4f.Provider.GptGo ✔️ ✔️ Unknown
gpttalk.ru g4f.Provider.GptTalkRu ✔️ ✔️ Unknown
koala.sh g4f.Provider.Koala ✔️ ✔️ Active
app.myshell.ai g4f.Provider.MyShell ✔️ ✔️ Unknown
onlinegpt.org g4f.Provider.OnlineGpt ✔️ ✔️ Unknown
perplexity.ai g4f.Provider.PerplexityAi ✔️ ✔️ Unknown
poe.com g4f.Provider.Poe ✔️ ✔️ Unknown ✔️
talkai.info g4f.Provider.TalkAi ✔️ ✔️ Unknown
you.com g4f.Provider.You ✔️ ✔️ Active
aitianhu.com g4f.Provider.AItianhu ✔️ ✔️ Inactive
e.aiask.me g4f.Provider.AiAsk ✔️ ✔️ Inactive
chatgpt.bestim.org g4f.Provider.Bestim ✔️ ✔️ Inactive
chatanywhere.cn g4f.Provider.ChatAnywhere ✔️ ✔️ Inactive
chatgpt4online.org g4f.Provider.Chatgpt4Online ✔️ ✔️ Inactive
chat.chatgptdemo.ai g4f.Provider.ChatgptDemoAi ✔️ ✔️ Inactive
chatgptfree.ai g4f.Provider.ChatgptFree ✔️ Inactive
chatgptlogin.ai g4f.Provider.ChatgptLogin ✔️ ✔️ Inactive
chatgptx.de g4f.Provider.ChatgptX ✔️ ✔️ Inactive
chat-shared2.zhile.io g4f.Provider.FakeGpt ✔️ ✔️ Inactive
freegpts1.aifree.site g4f.Provider.FreeGpt ✔️ ✔️ Inactive
gptgod.site g4f.Provider.GptGod ✔️ ✔️ Inactive
hashnode.com g4f.Provider.Hashnode ✔️ ✔️ Inactive
sdk.vercel.ai g4f.Provider.Vercel ✔️ ✔️ Inactive
chat.ylokh.xyz g4f.Provider.Ylokh ✔️ ✔️ Inactive

Other

Website Provider GPT-3.5 GPT-4 Stream Status Auth
bard.google.com g4f.Provider.Bard Unknown ✔️
deepinfra.com g4f.Provider.DeepInfra ✔️ Active
gemini.google.com g4f.Provider.Gemini Active ✔️
huggingface.co g4f.Provider.HuggingChat ✔️ Active
llama2.ai g4f.Provider.Llama2 ✔️ Active
labs.perplexity.ai g4f.Provider.PerplexityLabs ✔️ Active
phind.com g4f.Provider.Phind ✔️ Unknown
pi.ai g4f.Provider.Pi ✔️ Unknown
theb.ai g4f.Provider.ThebApi Unknown ✔️
open-assistant.io g4f.Provider.OpenAssistant ✔️ Inactive ✔️

Models

Model Base Provider Provider Website
gpt-3.5-turbo OpenAI 5+ Providers openai.com
gpt-4 OpenAI 2+ Providers openai.com
gpt-4-turbo OpenAI g4f.Provider.Bing openai.com
Llama-2-7b-chat-hf Huggingface 2+ Providers huggingface.co
Llama-2-13b-chat-hf Huggingface 2+ Providers huggingface.co
Llama-2-70b-chat-hf Huggingface 4+ Providers huggingface.co
CodeLlama-34b-Instruct-hf Huggingface 3+ Providers huggingface.co
Mixtral-8x7B-Instruct-v0.1 Huggingface 3+ Providers huggingface.co
Mistral-7B-Instruct-v0.1 Huggingface 3+ Providers huggingface.co
dolphin-2.6-mixtral-8x7b Huggingface g4f.Provider.DeepInfra huggingface.co
lzlv_70b_fp16_hf Huggingface g4f.Provider.DeepInfra huggingface.co
airoboros-70b Huggingface g4f.Provider.DeepInfra huggingface.co
airoboros-l2-70b-gpt4-1.4.1 Huggingface g4f.Provider.DeepInfra huggingface.co
openchat_3.5 Huggingface 2+ Providers huggingface.co
gemini Google g4f.Provider.Gemini gemini.google.com
gemini-pro Google 2+ Providers gemini.google.com
claude-v2 Anthropic 2+ Providers anthropic.com
pi Inflection g4f.Provider.Pi inflection.ai
🎁 Projects Stars 📚 Forks 🛎 Issues 📬 Pull requests
gpt4free Stars Forks Issues Pull Requests
gpt4free-ts Stars Forks Issues Pull Requests
Free AI API's & Potential Providers List Stars Forks Issues Pull Requests
ChatGPT-Clone Stars Forks Issues Pull Requests
ChatGpt Discord Bot Stars Forks Issues Pull Requests
Nyx-Bot (Discord) Stars Forks Issues Pull Requests
LangChain gpt4free Stars Forks Issues Pull Requests
ChatGpt Telegram Bot Stars Forks Issues Pull Requests
ChatGpt Line Bot Stars Forks Issues Pull Requests
Action Translate Readme Stars Forks Issues Pull Requests
Langchain Document GPT Stars Forks Issues Pull Requests
python-tgpt Stars Forks Issues Pull Requests

🤝 Contribute

Create Provider with AI Tool

Call in your terminal the create_provider.py script:

python etc/tool/create_provider.py
  1. Enter your name for the new provider.
  2. Copy and paste the cURL command from your browser developer tools.
  3. Let the AI create the provider for you.
  4. Customize the provider according to your needs.

Create Provider

  1. Check out the current list of potential providers, or find your own provider source!
  2. Create a new file in g4f/Provider with the name of the Provider
  3. Implement a class that extends BaseProvider.
from __future__ import annotations

from ..typing import AsyncResult, Messages
from .base_provider import AsyncGeneratorProvider

class HogeService(AsyncGeneratorProvider):
    url                   = "https://chat-gpt.com"
    working               = True
    supports_gpt_35_turbo = True

    @classmethod
    async def create_async_generator(
        cls,
        model: str,
        messages: Messages,
        proxy: str = None,
        **kwargs
    ) -> AsyncResult:
        yield ""
  1. Here, you can adjust the settings, for example, if the website does support streaming, set supports_stream to True...
  2. Write code to request the provider in create_async_generator and yield the response, even if it's a one-time response, do not hesitate to look at other providers for inspiration
  3. Add the Provider Name in g4f/Provider/__init__.py
from .HogeService import HogeService

__all__ = [
  HogeService,
]
  1. You are done !, test the provider by calling it:
import g4f

response = g4f.ChatCompletion.create(model='gpt-3.5-turbo', provider=g4f.Provider.PROVIDERNAME,
                                    messages=[{"role": "user", "content": "test"}], stream=g4f.Provider.PROVIDERNAME.supports_stream)

for message in response:
    print(message, flush=True, end='')

🙌 Contributors

A list of the contributors is available here
The Vercel.py file contains code from vercel-llm-api by @ading2210, which is licensed under the GNU GPL v3
Top 1 Contributor: @hlohaus

This program is licensed under the GNU GPL v3

xtekky/gpt4free: Copyright (C) 2023 xtekky

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program.  If not, see <https://www.gnu.org/licenses/>.

Star History

Star History Chart

📄 License


This project is licensed under GNU_GPL_v3.0.

(🔼 Back to top)