Thank you for confirming that the issue was resolved after the upgrade!
Regarding the delayed responses while using Google Gemini, could you please confirm whether you are using a paid Gemini API key or the free API tier?
The free API tier may have slower response times depending on the model, request complexity, and available API resources. Here are a few recommendations to improve response speed and accuracy:
- Expand Your Q&A Knowledge Base
Try adding as many relevant questions and answers as possible through Q&A Management. This gives the AI more structured information to reference, helping it provide more accurate and relevant responses.
- Configure RAG (Retrieval-Augmented Generation)
If you are using RAG, navigate to the Knowledge Retrieval Mode settings and check the available modes:
- RAG Only Mode: Uses retrieved knowledge to generate responses.
- Hybrid Mode: Combines retrieved knowledge with the AI model’s general knowledge.
You can try these modes to find what works best for your use case.
- Check RAG Processing Status
Navigate to RAG Management and review the total items and completed processing count.
If some items are still pending or have failed:
- Click Start Processing again to continue processing pending items.
- Retry failed items to ensure all your knowledge content is properly indexed.
- Consider Using a Paid Gemini / OpenAI API Key
Please note that RAG embedding and processing can be resource intensive. The free Gemini API quota may not be sufficient to complete all embedding operations, depending on the model and available limits.
Also, Content Indexing Mode may take longer because the system needs to process and index your content before it can be retrieved.
Once everything is properly indexed, try testing the responses again.
We hope these suggestions help improve both response speed and accuracy. If the issue persists, please let us know which Gemini model and API tier you are currently using so we can investigate further.
We appreciate your patience and are always happy to help!