Send replies to new Slack channel messages with ChatGPT conversations routed between strong and weak OpenAI models using Not Diamond
Set up this Zap to build a Slack chatbot that saves on cost without compromising quality. This integration uses Not Diamond to automatically route each new message in a specified Slack channel to the ideal OpenAI model—selecting between a stronger, high-quality model (GPT-4o) on more complex messages and a weaker cost-effective option (GPT-4o mini) on simpler messages. The chosen model generates a response that’s instantly posted back in the channel, enhancing team communication with optimized efficiency.
Set up this Zap to build a Slack chatbot that saves on cost without compromising quality. This integration uses Not Diamond to automatically route each new message in a specified Slack channel to the ideal OpenAI model—selecting between a stronger, high-quality model (GPT-4o) on more complex messages and a weaker cost-effective option (GPT-4o mini) on simpler messages. The chosen model generates a response that’s instantly posted back in the channel, enhancing team communication with optimized efficiency.
- When this happens...New Message Posted to Channel
Triggers when a new message is posted to a specific #channel you choose.
- automatically do this...Route Prompt
Sends the prompt to the /modelSelect endpoint which will return a label for which LLM should be called.
- then do this...Conversation (Legacy)
Sends a Chat to OpenAI and generates a Completion, optionally storing messages as we do. Powered by Chat Completions API.
- then do this!Send Channel Message
Triggers when a new message is posted to a specific #channel you choose.
- Free forever for core features
- 14 day trial for premium features & apps