Learn to say popular text embedding model names correctly.
0 / 5 completed
1 / 5
How is text-embedding-3 (OpenAI's third-generation family of text embedding models) correctly pronounced?
text-embedding-3 is pronounced 'TEKST-em-BED-ing THREE' — 'text embedding' plus the number 'three'. In a technical interview: "text-embedding-3 gave us noticeably better retrieval accuracy than the previous model version."
2 / 5
How is Cohere Embed (Cohere's multilingual text embedding model family) correctly pronounced?
Cohere Embed is pronounced 'koh-HEER em-BED' — 'cohere' stress on HEER, plus 'embed'. In a technical interview: "Cohere Embed handled the French support tickets just as well as the English ones."
3 / 5
How is BGE (BAAI General Embedding, a widely used open-source embedding model family) correctly pronounced?
BGE is pronounced 'BEE-JEE-EE' — every letter spoken individually, B-G-E. In a technical interview: "BGE ranked the relevant passage first in nearly every one of our retrieval test cases."
4 / 5
How is E5 (Microsoft's family of general-purpose text embedding models) correctly pronounced?
E5 is pronounced 'EE-FYV' — the letter 'E' plus the number 'five'. In a technical interview: "E5 embedded both the query and the documents into the same vector space in one pass."
5 / 5
How is Instructor-XL (instruction-tuned embedding model that adapts its output to a stated task) correctly pronounced?
Instructor-XL is pronounced 'in-STRUK-ter EKS-EL' — 'instructor' exactly like the everyday word for a teacher, plus 'X-L' spoken as letters. In a technical interview: "Instructor-XL produced a different embedding for the same sentence depending on whether we asked it to classify or to search."