Titan Text Embeddings V2
Titan Text Embeddings V2 is the current established Titan text-vector release. Bedrock lists it as Active.
What the release provides
The API model ID is amazon.titan-embed-text-v2:0. It returns embeddings for retrieval rather than a generated answer.Titan Text Embeddings V2 Bedrock model card
How to use it well
Practical guidance. Choose document sections that preserve meaning and return their source locations with each match. Review cross-language queries separately from same-language search. A high similarity score is a retrieval signal, not proof that a passage answers the question.
Input length and vector dimensions
The guide allows up to 8,192 tokens or 50,000 characters. Output dimensions are 1,024 by default, with 512 and 256 also available. It is optimized for English; AWS identifies cross-language retrieval as weaker than same-language retrieval.Titan text embedding specifications
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API |
| Model identifiers | amazon.titan-embed-text-v2:0 |