#OpenAI Chat Discord Bot

14 messages · Page 1 of 1 (latest)

gray seal
#

I created a discord bot to create responses with the GPT-3.5 turbo model.
It can generate single responses or it can create a thread where you can create followup requests (like the chat.openai.com Web-UI)

It is the first version and will be massively improved in the future

https://github.com/ZeldaFan0225/ChatGPT-Discord-Bot

GitHub

A Discord bot interacting with OpenAIs GPT-3.5 model - GitHub - ZeldaFan0225/ChatGPT-Discord-Bot: A Discord bot interacting with OpenAIs GPT-3.5 model

lost kestrelBOT
#

@gray seal

Zelda_Fan's message blocked by AutoMod

We are committed to fostering a positive atmosphere and have implemented additional filters to block any inappropriate or disruptive language and topics.

gray seal
#

sorry

#

GPT-3.5 turbo Discord Bot

sage tiger
#

that’s sick

gray seal
#

It's a very early version though

gray seal
#

Updated the bot to have selectable system instructions for fun responses, and more

onyx seal
#

I feel zelda doesn't show their own work off enough, I've been using this for a while - it's great.

The many options for system instructions, selectable 3.5 or 4.0 (assuming you have API access). Ability to start threads off single, the newly added "hey gpt" type option which even grabs the user's name, really don't give yourself enough credit. Works great.

onyx seal
#

bumping as zelda updated to support the new models already 🙂

gray seal
#

well on that note i added the new models, adjusted priving and some feature which were added after i made this post are:

  • leaderboard to see per user usage (with money spent per user)
  • blacklist if a user misbehaves
  • a tokenizer to show the tokens gpt generates
  • configurable context actions that you give a system instruction and chatgpt gets the messages content
  • selectable system instructions
  • moderation (all prompts can be run through the moderation endpoint)
gray seal
#

OpenAI Chat Discord Bot

#

Now with support for non openai sources like self hosted llama for direct comparison

#

Also with support for all currently available models offered by openai (depending on which token is used access might be restricted)