spring-projects / spring-projects/spring-ai
AWS Bedrock Cohere Embedding Model returns no metrics
Nobody has claimed this yet.
- Dominant language
- Java
- Stars
- 9.5k
- Forks
- 2.9k
- Avg merge
- 1d 10h
- Merged PRs (30d)
- 5
Description
Bug description
When using the AWS Bedrock Cohere Embedding Model (cohere.embed-multilingual-v3), no metrics are included in the response. So I cannot track the used tokens.
When logging the raw response, the input token usage is included as HTTP response header, so Spring AI could parse it and fill the metrics.
e.g. "X-Amzn-Bedrock-Input-Token-Count: 743", "X-Amzn-Bedrock-Invocation-Latency: 186".
Environment
Java 21, Spring AI 1.1.6
Steps to reproduce
Configure the AWS Bedrock Cohere Embedding Model.
Submit any request.
See that no metrics are attached.
Expected behavior
The metrics were set from the information included in the HTTP headers.
Setup
Application Properties:
spring.ai.model.embedding=bedrock-cohere
spring.ai.bedrock.cohere.embedding.model=cohere.embed-multilingual-v3
spring.ai.bedrock.cohere.embedding.options.input-type=CLUSTERING
Service:
import org.springframework.ai.embedding.EmbeddingModel;
import org.springframework.ai.embedding.EmbeddingRequest;
import org.springframework.ai.embedding.EmbeddingResponse;
import org.springframework.stereotype.Service;
import lombok.RequiredArgsConstructor;
@Service
@RequiredArgsConstructor
class EmbeddingServiceImpl implements EmbeddingService {
private final EmbeddingModel model;
@Override
public EmbeddingResponse embed(EmbeddingRequest request) {
EmbeddingResponse response = embeddingModel.call(embeddingRequest); // <-- call Spring AI
return response;
}
}
Dependencies:
dependencies {
implementation("org.springframework.ai:spring-ai-starter-model-bedrock-converse")
implementation("org.springframework.ai:spring-ai-starter-model-bedrock")
implementation("software.amazon.awssdk:apache-client")
}
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by tracing the AWS Bedrock Cohere Embedding Model from the EmbeddingModel call and inspect how the raw HTTP response headers are handled. Reproduce the request with cohere.embed-multilingual-v3 and verify that X-Amzn-Bedrock-Input-Token-Count and X-Amzn-Bedrock-Invocation-Latency are reflected in the returned metrics.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- aws, java, spring
- Domain
- ai, backend
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 55/100