The best embedding model for RAG is the one that retrieves relevant evidence from your corpus within your latency and indexing budget. Start with text-embedding-3-small or large for a hosted text base
Markdown supported · ⌘↵ to post
No responses yet.