1
0
Fork
You've already forked ChatGPT-to-API
0
No description
  • Go 95.8%
  • Dockerfile 1.6%
  • Python 1.4%
  • Shell 1.2%
Find a file
2023年08月08日 22:14:02 +08:00
.github/workflows Create release.yml 2023年06月26日 08:50:56 +00:00
cases/sider Add a example ~sider 2023年05月29日 14:09:08 +08:00
conversion fix GPT-4 model, structural change of access_tokens.json, add har file tutorial to readme 2023年08月04日 09:46:00 +08:00
docs fix: 404 Error 2023年07月11日 13:59:00 +08:00
internal Merge pull request #110 from oldweipro/fix-403 2023年08月07日 22:38:09 +08:00
tools fix 403 2023年08月07日 11:31:03 +08:00
typings feat response fmt like openai api 2023年06月20日 15:24:37 +08:00
.env.example feat: support PORT env 2023年06月28日 11:58:54 +08:00
.gitignore fix GPT-4 model, structural change of access_tokens.json, add har file tutorial to readme 2023年08月04日 09:46:00 +08:00
auth.go fix GPT-4 model, structural change of access_tokens.json, add har file tutorial to readme 2023年08月04日 09:46:00 +08:00
docker-compose.yml Update docker-compose.yml 2023年07月07日 15:01:34 +00:00
Dockerfile Update Dockerfile 2023年04月12日 11:29:45 +08:00
go.mod fix GPT-4 model, structural change of access_tokens.json, add har file tutorial to readme 2023年08月04日 09:46:00 +08:00
go.sum fix GPT-4 model, structural change of access_tokens.json, add har file tutorial to readme 2023年08月04日 09:46:00 +08:00
handlers.go fix GPT-4 model, structural change of access_tokens.json, add har file tutorial to readme 2023年08月04日 09:46:00 +08:00
main.go fix GPT-4 model, structural change of access_tokens.json, add har file tutorial to readme 2023年08月04日 09:46:00 +08:00
middleware.go more verbose 2023年06月17日 11:37:39 +08:00
README.md fix readme url 2023年08月08日 12:26:48 +08:00
README_CN.md Update README_CN.md 2023年07月07日 05:24:26 +00:00
README_JA.md fix readme url 2023年08月08日 12:26:48 +08:00
README_ZH.md fix GPT-4 model, structural change of access_tokens.json, add har file tutorial to readme 2023年08月04日 09:46:00 +08:00

ChatGPT-to-API

Create a fake API using ChatGPT's website

IMPORTANT

You will not get free support for this repository. This was made for my own personal use and documentation will continue to be limited as I don't really need documentation. You will find more detailed documentation in the Chinese docs by a contributor.

API endpoint: http://127.0.0.1:8080/v1/chat/completions.

中文文档(Chinese Docs)

Setup

Authentication

Access token and PUID(only for PLUS account) retrieval has been automated by OpenAIAuth with account email & password.

accounts.txt - A list of accounts separated by new line

Format:

email:password
...

All authenticated access tokens and PUID will store in access_tokens.json

Auto renew access tokens and PUID after 7 days

Caution! please use unblocked ip for authentication, first login to https://chat.openai.com/ to check ip availability if you can.

GPT-4 Model (Optional)

If you configured a PLUS account and use the GPT-4 model, a HAR file (chat.openai.com.har) is required to complete CAPTCHA verification

  1. Use a chromium-based browser (Chrome, Edge) or Safari to login to https://chat.openai.com/, then open the browser developer tools (F12), and switch to the Network tab.

  2. Create a new chat and select the GPT-4 model, ask a question at will, click the Export HAR button under the Network tab, export the file chat.openai.com.har

API Authentication (Optional)

Custom API keys for this fake API, just like OpenAI api

api_keys.txt - A list of API keys separated by new line

Format:

sk-123456
88888888
...

Getting set up

git clone https://github.com/acheong08/ChatGPT-to-API
cd ChatGPT-to-API
go build
./freechatgpt

Environment variables

  • PUID - A cookie found on chat.openai.com for Plus users. This gets around Cloudflare rate limits
  • SERVER_HOST - Set to 127.0.0.1 by default
  • SERVER_PORT - Set to 8080 by default
  • ENABLE_HISTORY - Set to true by default

Files (Optional)

  • proxies.txt - A list of proxies separated by new line

    http://127.0.0.1:8888
...

  • access_tokens.json - A JSON array of access tokens for cycling (Alternatively, send a PATCH request to the correct endpoint)

    [{token:"access_token1", puid:"puid1"}, {token:"access_token2", puid:"puid2"}...]

Admin API docs

https://github.com/acheong08/ChatGPT-to-API/blob/master/docs/admin.md

API usage docs

https://platform.openai.com/docs/api-reference/chat