Hey Bubblers !
Did you know that your model and system prompt could be exposed to users when using text-based streaming plugins?
This is a serious concern, as it opens the door for unintended access to sensitive or proprietary information, potentially compromising the design of your solution or product.
Imagine:
-
Your finely tuned prompt – painstakingly crafted for optimal performance – suddenly exposed.
-
Your model’s identity – the backbone of your service – laid bare, making it easier for competitors to replicate or exploit your work.
That’s not just risky; it’s avoidable.
Here’s the Solution
All the following text-based streaming plugins from wise:able now support system prompt and model hiding from your users!
- OpenAI - ChatGPT Streaming
- Azure AI - ChatGPT on your Data
- Mistral AI - Chat Streaming
- Google AI - Gemini Chat Streaming
- Claude AI - Chat Streaming
- Groq - Lightspeed AI Chat Streaming
- Perplexity AI - Chat Streaming
Enjoy!