Mar 4 • General discussion 💬
VAPI Knowledgebase Update
VAPI added a Knowledgebase feature, now accessible via the Files section in the UI! This utilizes Gemini 1.5 Flash through a tool, and based on their observations, it delivers significantly better retrieval-augmented generation (RAG) quality.
How to Use It:
✅Upload files in the Files section.
✅Attach them to a section and provide a relevant prompt.
✅Instruct the model to call QueryTool when you need to fetch details.
🧠Things to Keep in Mind:
More context = higher latency
Max limit: 1M tokens (~1500 pages) (Recommended: 50-100 pages for optimal performance.)
Give it a try and share your feedback!
6
2 comments
VAPI Knowledgebase Update
skool.com/brendan
Learn To Make Money With AI!
- 50+ Free AI Agent Templates
- 60+ Free AI Course Videos
(n8n, Make, Vapi, Voiceflow)
- AI Software Discounts 💰
Leaderboard (30-day)
1
+76
2
+71
3
+60
4
+58
5
+58
Powered by