Selecting an embedding model

3.1Standard

Build multilingual document search.

Balance semantic quality, language coverage, latency, and cost.

Choose two selection criteria.

Hint 1: Focus on the requirement

Select embeddings with language and domain evaluation, and include storage, latency, and re-embedding work in the lifecycle decision.

Hint 2: Key terms

Review embedding-model selection and re-embedding.

Embedding model: A model that converts content into vectors intended to preserve useful semantic relationships.
Re-embedding: Recomputing stored vectors when the embedding model or source documents change.
Vector dimension: The number of numeric components in each embedding vector.
Domain evaluation: Testing a model on representative data from the language and subject area it will serve.

Question IDAWS-AIF-301

EnvironmentAWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認

Last technical review2026-08-20

More practice in this area