GPT API free (free API Key limited to 200 requests/day)

GPT API free program features

1. Support Models, Embedding, text-davinci(not supported by free version), GPT-3.5-Turbo, GPT-3.5-Turbo-16K(not supported by free version), GPT-4, DALLE(not supported by free version), Whisper(not supported by free version). (AutoGPT, gpt_academic, langchain, etc. are supported by the free version).
2. The free version supports gpt-4, 3 times a day; it supports gpt-4o-mini and shares 200 times a day with gpt-3.5-turbo.
3. Fully consistent with the official interface standard, compatible with a variety of software/plug-ins.
4. Streaming responses are supported.
5. Domestic lines use dynamic acceleration, and the experience is far superior to using a proxy to connect to the official.
6. No need for scientific internet access, directly available in domestic environment.
7. Completely free for personal use.

Free API Key may only be used for personal, non-commercial use, education, and non-profit scientific research. Free API Key is strictly prohibited for commercial use, and large-scale training of commercial models is strictly prohibited!

Free API Key limit 200 requests/day/IP & Key call frequency (gpt and embedding are counted separately, each 200 times), that is to say, if you use more than one Key under one IP, the total number of requests of all Keys per day can not be more than 200; similarly, if you use a Key for more than one IP, the number of requests of this Key per day can not be more than 200. (Paid API does not have this limit. Similarly, if you use a Key for multiple IPs, the number of requests per day for this Key cannot exceed 200.(The paid version of the API does not have this limitation.)

Instructions for use

Project Address

🚀 Apply to receive a free API Key for internal testing
The free version supports gpt-3.5-turbo, embedding, gpt-4o-mini, gpt-4. Among them, gpt-4 is limited to 3 calls per day due to its high price (0:00 refresh). If you need more stable and fast gpt-4, please use the paid version.
The free version of gpt-4 is served by gpt-4o, but the free version does not support map reading at this time.
Forwarding Host1: https://api.chatanywhere.tech (domestic transit, lower latency)
Forwarding Host2: https://api.chatanywhere.cn (for foreign use, domestic need global proxy)
Balance and Usage Inquiry (notices are also posted here). Balance Enquiry and Announcement
Forwarding API can not directly to the official interface api.openai.com to initiate a request, you need to change the request address to api.chatanywhere.tech to use, most plug-ins and software can be modified.

This API Key is used for forwarding API, you need to change Host to api.chatanywhere.tech (preferred in China) or api.chatanywhere.cn (used in foreign countries, global proxy is needed in China).

code example

python openai official library (using AutoGPT, langchain, etc.)

Method 1

from openai import OpenAI

client = OpenAI(
# defaults to os.environ.get("OPENAI_API_KEY")
api_key="YOUR API KEY",
base_url="https://api.chatanywhere.tech/v1"
# base_url="https://api.chatanywhere.cn/v1"
)

Method 2

Modify the environment variable OPENAI_API_BASE, how to change the environment variable of each system, please search by yourself, please restart the system after modifying the environment variable does not work.

OPENAI_API_BASE=https://api.chatanywhere.tech/v1
或 OPENAI_API_BASE=https://api.chatanywhere.cn/v1

Open source gpt_academic

Find the API_URL_REDIRECT configuration in the config.py file and modify it to the following:

API_URL_REDIRECT = {"https://api.openai.com/v1/chat/completions": "https://api.chatanywhere.tech/v1/chat/completions"}
# API_URL_REDIRECT = {"https://api.openai.com/v1/chat/completions": "https://api.chatanywhere.cn/v1/chat/completions"}

BotGem (AMA)

ChatGPT desktop app, support all platforms, support gpt-4-vision. how to use: download and install it, set it in settings as shown in the picture and click update.