Mintora Product Notes
How Mintora AI Generation Works
A transparent overview of the current AI generation flow used by Mintora.
1. You submit a prompt
Mintora's AI generation endpoint accepts a prompt, an optional system instruction and a tool type. The endpoint validates the request before sending it to the configured AI provider.
2. The request is sent to PublicAI
The current backend implementation routes this generation endpoint through the PublicAI API. The model name is configured on the server, so the exact model can change with the platform configuration.
3. Mintora returns the response
When the provider returns a response, Mintora sends the generated text back to the requesting website feature. The endpoint supports optional account authentication and applies request rate limiting.
4. Generation history may be saved
After generation, the backend attempts to save the submitted prompt, generated response and related provider/model usage information. History persistence is best-effort: a database history write failure does not by itself cancel an otherwise successful generation.
Use AI output thoughtfully
AI responses can contain mistakes or omit important context. Review and verify important information before using it, especially for legal, medical, financial or other consequential decisions. Avoid entering passwords, payment credentials or other sensitive information in prompts.
For details about prompt handling and other information processed by the platform, read the Mintora Privacy Policy.