2024-01-18 10:58:04 +10:00
2023-11-10 06:41:57 +10:00
2023-11-10 07:57:50 +10:00
2024-01-17 04:12:38 +10:00
2023-12-23 14:00:28 +10:00
2024-01-17 04:32:27 +10:00
2023-12-14 08:29:35 +10:00
2023-12-17 09:15:56 +10:00
2024-01-18 10:58:04 +10:00
2024-01-18 10:34:00 +10:00
2023-12-11 03:44:49 +10:00

keychecker

a bulk key checker for various AI services

Currently supports and validates keys for the services below, and checks for the listed attributes a key might have:

  • OpenAI - (Best model, key in quota, RPM (catches increase requests), tier, list of organizations if applicable, trial key status)
  • Anthropic - (Pozzed check)
  • AI21 - (Trial check)
  • Google MakerSuite (Gemini)
  • AWS - (Admin status, auto-fetch the region, logging status, username, bedrock status)
  • Azure - (Auto-fetch all deployments, auto-fetch best deployment/model, filter status)
  • Google Cloud Vertex AI - (Requires a key file since oauth tokens expire hourly. Good luck scraping for those.)
  • MistralAI - (Subscription status)

An asynchronus version of the checker can be found in the branch 'async'. I don't want to merge into main just yet because of the issues OAI had recently when speedchecking lots of keys. It is currently working though and can process around 1.5k keys every ~2 seconds.

Usage:

pip install -r requirements.txt

python main.py

Optional Arguments:

-proxyoutput

Outputs keys in a format that can be easily copy pasted into khanon's proxy instead of pretty print

-nooutput

Stops outputting and saving keys to the snapshot file (proxyoutput will also do this)

-file

Reads keys from a file instead of stdin, place either the absolute or relative path to the file in quotes after the flag.

S
Description
a key checker for various AI services
Readme MIT 313 KiB
Languages
Python 100%