llama-index-embeddings-bedrock

What it is and what it does

This package integrates Amazon Bedrock embedding models with the embedding interface, letting you generate text embeddings using AWS-hosted models without managing your own embedding infrastructure. It supports multiple Bedrock foundation models—Amazon Titan variants and Cohere v3/v4—and handles the API differences between them transparently, including Cohere v4's multimodal capabilities and response format changes.

You configure it with a model name and AWS region, then call `get_text_embedding()` or `get_text_embedding_batch()` to convert text into vectors. AWS credentials flow through boto3's standard chain (environment variables, IAM roles, profiles, or explicit parameters). The integration also supports Application Inference Profiles for cost tracking and model usage governance in enterprise Bedrock deployments.

Use it for:

• Embed documents in a RAG pipeline using Bedrock models without running your own embedding server.
• Build semantic search over text corpora by generating embeddings via Cohere or Titan through the integration.
• Use Cohere v4 multimodal embeddings to embed both text and images in a single unified vector space.
• Integrate Bedrock embeddings with automatic response format handling across model versions.
• Track embedding costs and usage through AWS Application Inference Profiles in governed environments.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Provides Amazon Bedrock embedding models integration for LlamaIndex, enabling text-to-vector conversion using Titan and Cohere embeddings through AWS.

Yes. Low install friction, active maintenance, no known vulnerabilities, and permissive MIT license. Install if you need Bedrock embeddings and have AWS access. Skip if you prefer self-hosted embeddings or use a different embedding provider.

Install

llama-index-embeddings-bedrock on PyPI

pip install llama-index-embeddings-bedrock

uv add llama-index-embeddings-bedrock

poetry add llama-index-embeddings-bedrock

Installing llama-index-embeddings-bedrock

Before you install

Low install friction with a pure-Python wheel. Actively maintained as of 2026-08-14. Requires boto3 and llama-index-core as runtime dependencies.

License in practice

MIT license permits commercial and private use with minimal restrictions, making this suitable for most production deployments.

Quickstart

pip install llama-index-embeddings-bedrock

from llama_index.embeddings.bedrock import BedrockEmbedding

embed_model = BedrockEmbedding(
    model_name="cohere.embed-english-v3",
    region_name="us-east-1",
)
embedding = embed_model.get_text_embedding("Hello world")

Requires AWS credentials (via environment variables, profile, or direct parameters) and access to Amazon Bedrock in your AWS account.

Verify before relying

• Performance characteristics and latency for batch embedding operations at scale.
• Cost implications of different Bedrock models and request volumes.
• Whether Application Inference Profile ARN mismatches produce silent failures or explicit errors.

Package facts

Evidence: llama_index_embeddings_bedrock-0.8.3-py3-none-any.whl

Tags

More Artificial Intelligence packages