Amazon Bedrock Chat and Embedding Models

Amazon Bedrock is a managed service that provides access to foundation models from AWS and other providers. The Quarkus LangChain4j extension integrates its chat, streaming chat, and embedding models.

Prerequisites

Before configuring the extension:

  • Enable access to the model you intend to use in Amazon Bedrock.

  • Choose an AWS Region where that model is available.

For the models and Regions supported by Bedrock, see the AWS model catalog.

Dependency

Add the following dependency to your project:

<dependency>
  <groupId>io.quarkiverse.langchain4j</groupId>
  <artifactId>quarkus-langchain4j-bedrock</artifactId>
  <version>1.15.0.CR1</version>
</dependency>

Even better, if you use the Quarkus platform BOM (default for projects generated), add the Quarkus Langchain4J BOM and all dependency versions will align:

    <dependencyManagement>
        <dependencies>
            <dependency>
                <groupId>${quarkus.platform.group-id}</groupId>
                <artifactId>${quarkus.platform.artifact-id}</artifactId>
                <version>${quarkus.platform.version}</version>
                <type>pom</type>
                <scope>import</scope>
            </dependency>
            <dependency>
                <groupId>${quarkus.platform.group-id}</groupId>
                <artifactId>quarkus-langchain4j-bom</artifactId> (1)
                <version>${quarkus.platform.version}</version> (2)
                <type>pom</type>
                <scope>import</scope>
            </dependency>
        </dependencies>
    </dependencyManagement>

    <dependencies>
      <dependency>
        <groupId>io.quarkiverse.langchain4j</groupId>
        <artifactId>quarkus-langchain4j-bedrock</artifactId>
        (3)
      </dependency>
    </dependencies>
1 In your dependencyManagement section, add the quarkus-langchain4j-bom
2 Inherit the version from your platform version
3 Voilà, no need for version alignment anymore

If no other chat model extension is selected, AI Services will automatically use the configured Bedrock chat model.

AWS Credentials

The extension uses the AWS SDK default credentials provider chain. For local development, configure an AWS profile or set the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY environment variables.

When running on AWS, credentials can be supplied by the runtime, such as ECS task roles, EC2 instance profiles, or web identity tokens on EKS. Do not configure AWS credentials in application.properties.

For the complete list and order of credential sources, see the AWS SDK default credentials provider chain documentation.

Chat Model

Configure the Region and a Bedrock model ID in application.properties:

quarkus.langchain4j.bedrock.aws.region=us-east-1
quarkus.langchain4j.bedrock.chat-model.model-id=global.amazon.nova-2-lite-v1:0

# Optional generation parameters
quarkus.langchain4j.bedrock.chat-model.temperature=0.2
quarkus.langchain4j.bedrock.chat-model.max-tokens=1024

The extension registers both ChatModel and StreamingChatModel beans. You can inject them directly:

import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.chat.StreamingChatModel;
import jakarta.enterprise.context.ApplicationScoped;

@ApplicationScoped
class AssistantService {

    private final ChatModel chatModel;
    private final StreamingChatModel streamingChatModel;

    AssistantService(ChatModel chatModel, StreamingChatModel streamingChatModel) {
        this.chatModel = chatModel;
        this.streamingChatModel = streamingChatModel;
    }
}

Or, declare an AI Service and let Quarkus inject the configured model:

import dev.langchain4j.service.UserMessage;
import io.quarkiverse.langchain4j.RegisterAiService;
import io.smallrye.mutiny.Multi;

@RegisterAiService
public interface Assistant {
    String chat(String message);

    @UserMessage("Answer the question: {message}")
    Multi<String> stream(String message);
}

When an AI Service method returns Multi<String>, Quarkus uses the configured StreamingChatModel. Each item emitted by the Multi is part of the response. For REST and WebSocket examples, see the streamed responses guide.

AI Services are request-scoped by default. Apply a CDI scope when your chat memory lifecycle requires it.

Named Configurations

Use named configurations when an application needs more than one Bedrock model. Select a named model in an AI service or inject it with @ModelName.

quarkus.langchain4j.bedrock."fast".aws.region=us-east-1
quarkus.langchain4j.bedrock."fast".chat-model.model-id=global.amazon.nova-2-lite-v1:0
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.chat.StreamingChatModel;
import io.quarkiverse.langchain4j.ModelName;
import jakarta.enterprise.context.ApplicationScoped;

@ApplicationScoped
class AssistantService {

    private final ChatModel chatModel;
    private final StreamingChatModel streamingChatModel;

    AssistantService(@ModelName("fast") ChatModel chatModel,
            @ModelName("fast") StreamingChatModel streamingChatModel) {
        this.chatModel = chatModel;
        this.streamingChatModel = streamingChatModel;
    }
}

Embedding Model

The extension also provides an EmbeddingModel. Configure it independently from the chat model:

quarkus.langchain4j.bedrock.aws.region=us-east-1
quarkus.langchain4j.bedrock.embedding-model.model-id=cohere.embed-english-v3

For Amazon Titan embedding models, use the embedding-model.titan configuration group. For Cohere embedding models, use embedding-model.cohere; the configuration reference lists the provider-specific options.

You can inject the configured EmbeddingModel directly:

import dev.langchain4j.model.embedding.EmbeddingModel;
import jakarta.enterprise.context.ApplicationScoped;

@ApplicationScoped
class EmbeddingService {

    private final EmbeddingModel embeddingModel;

    EmbeddingService(EmbeddingModel embeddingModel) {
        this.embeddingModel = embeddingModel;
    }
}

Bedrock Features

The chat model supports Bedrock guardrails, prompt caching, and reasoning when the selected model supports them. Verify model support and regional availability before enabling these capabilities.

Guardrails

Configure an existing Amazon Bedrock guardrail with its identifier and version:

quarkus.langchain4j.bedrock.chat-model.guardrail.guardrail-identifier=<guardrail-id>
quarkus.langchain4j.bedrock.chat-model.guardrail.guardrail-version=<guardrail-version>

# Optional for StreamingChatModel: sync or async
quarkus.langchain4j.bedrock.chat-model.guardrail.stream-processing-mode=sync

For more information about creating and configuring guardrails, see the Amazon Bedrock Guardrails documentation.

Prompt caching

Prompt caching reduces latency and input token costs when a prompt contains repeated content. The following example uses Claude Opus 4.7, which supports explicit prompt caching and a one-hour cache TTL:

quarkus.langchain4j.bedrock.chat-model.model-id=us.anthropic.claude-opus-4-7
quarkus.langchain4j.bedrock.chat-model.prompt-caching=AFTER_SYSTEM
quarkus.langchain4j.bedrock.chat-model.prompt-caching-ttl=1h

prompt-caching-ttl applies only when prompt-caching is configured. The available cache placement options are AFTER_SYSTEM, AFTER_USER_MESSAGE, and AFTER_TOOLS.

For supported models, regions, token requirements, and TTL options, see the Amazon Bedrock prompt caching documentation.

Reasoning

For models that support adaptive reasoning, configure the reasoning effort. The following example uses Claude Opus 4.7:

quarkus.langchain4j.bedrock.chat-model.model-id=us.anthropic.claude-opus-4-7
quarkus.langchain4j.bedrock.chat-model.reasoning-effort=medium

For models that use a token budget instead of adaptive reasoning, configure reasoning:

quarkus.langchain4j.bedrock.chat-model.reasoning=1024

Do not configure reasoning and reasoning-effort together. When both are set, reasoning-effort takes precedence.

For model-specific reasoning support, see the Amazon Bedrock reasoning documentation.

Configuration Reference

Configuration property fixed at build time - All other configuration properties are overridable at runtime

Configuration property

Type

Default

Whether the model should be enabled

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_ENABLED

boolean

true

Whether the model should be enabled

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_ENABLED

boolean

true

Whether to enable the integration. Defaults to true, which means requests are made to Bedrock. Set to false to disable all requests.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_ENABLE_INTEGRATION

boolean

true

Whether the Bedrock client should log requests

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_LOG_REQUESTS

boolean

false

Whether the Bedrock client should log responses

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_LOG_RESPONSES

boolean

false

Whether chat model body in request and response should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_LOG_BODY

boolean

false

Region used by the bedrock runtime api. See Regions Supported.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_AWS_REGION

string

default aws region provider chain

Override the endpoint used by the bedrock client

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_AWS_ENDPOINT_OVERRIDE

string

Specify a custom credentials provider to use for the bedrock client. Identified by bean name.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_AWS_CREDENTIALS_PROVIDER

string

default aws credentials provider chain

The maximum number retries the aws sdk client will attempt.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_AWS_MAX_RETRIES

int

3

Configure the amount of time to allow the client to complete the execution of an API call. This timeout covers the entire client execution except for marshalling. This includes request handler execution, all HTTP requests including retries, unmarshalling, etc. This value should always be positive, if present.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_AWS_API_CALL_TIMEOUT

Duration 

13s

Connect Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_CONNECT_TIMEOUT

Duration 

3s

Read Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_TIMEOUT

Duration 

10s

The name of the proxy configuration to use, as registered in the Quarkus Proxy Registry.

Precedence rules:

  • If set, it takes precedence over the deprecated proxy-address, proxy-user, proxy-password and non-proxy-hosts properties, which are ignored.

  • If not set and a default proxy configuration is configured (quarkus.proxy.*), then that one is used.

  • If set, the configuration from quarkus.proxy.<name>.* is used.

  • If set but no proxy configuration is found with that name, an error is thrown at runtime.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_PROXY_CONFIGURATION_NAME

string

If true, the REST clients will not provide additional contextual information (like REST client class and method names) when exception occurs during a client invocation.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_DISABLE_CONTEXTUAL_ERROR_MESSAGES

boolean

false

The time in ms for which a connection remains unused in the connection pool before being evicted and closed. A timeout of 0 means there is no timeout.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_CONNECTION_TTL

int

The size of the connection pool for this client.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_CONNECTION_POOL_SIZE

int

If set to false disables the keep alive completely.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_KEEP_ALIVE_ENABLED

boolean

true

The class name of the host name verifier. The class must have a public no-argument constructor.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_HOSTNAME_VERIFIER

string

Set whether hostname verification is enabled. Default is enabled. This setting should not be disabled in production as it makes the client vulnerable to MITM attacks.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_VERIFY_HOST

boolean

The trust store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_TRUST_STORE

string

The trust store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_TRUST_STORE_PASSWORD

string

The type of the trust store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_TRUST_STORE_TYPE

string

The key store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_KEY_STORE

string

The key store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_KEY_STORE_PASSWORD

string

The type of the key store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_KEY_STORE_TYPE

string

The name of the TLS configuration to use.

If not set and the default TLS configuration is configured (quarkus.tls.*) then that will be used. If a name is configured, it uses the configuration from quarkus.tls.<name>.* If a name is configured, but no TLS configuration is found with that name then an error will be thrown.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_TLS_CONFIGURATION_NAME

string

Whether chat model requests should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_LOG_REQUESTS

boolean

false

Whether chat model responses should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_LOG_RESPONSES

boolean

false

Whether chat model body in request and response should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_LOG_BODY

boolean

false

Region used by the bedrock runtime api. See Regions Supported.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_AWS_REGION

string

default aws region provider chain

Override the endpoint used by the bedrock client

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_AWS_ENDPOINT_OVERRIDE

string

Specify a custom credentials provider to use for the bedrock client. Identified by bean name.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_AWS_CREDENTIALS_PROVIDER

string

default aws credentials provider chain

The maximum number retries the aws sdk client will attempt.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_AWS_MAX_RETRIES

int

3

Configure the amount of time to allow the client to complete the execution of an API call. This timeout covers the entire client execution except for marshalling. This includes request handler execution, all HTTP requests including retries, unmarshalling, etc. This value should always be positive, if present.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_AWS_API_CALL_TIMEOUT

Duration 

13s

Model id to use. See Models Supported

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_MODEL_ID

string

global.amazon.nova-2-lite-v1:0

What sampling temperature to use, between 0.0 and 1.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.

It is generally recommended to set this or the top-k property but not both.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_TEMPERATURE

double

0.7

The maximum number of tokens to generate in the completion.

The token count of your prompt plus max_tokens cannot exceed the model’s context length

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_MAX_TOKENS

int

1024

Double (0.0-1.0). Nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

It is generally recommended to set this or the temperature property but not both.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_TOP_P

double

1.0

Reduces the probability of generating nonsense. A higher value (e.g. 100) will give more diverse answers, while a lower value (e.g. 10) will be more conservative

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_TOP_K

int

The custom text sequences that will cause the model to stop generating

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_STOP_SEQUENCES

list of string

Specifies the desired format for the model’s output.

Allowable values: [text, json]

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_RESPONSE_FORMAT

string

Connect Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_CONNECT_TIMEOUT

Duration 

3s

Read Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_TIMEOUT

Duration 

10s

The name of the proxy configuration to use, as registered in the Quarkus Proxy Registry.

Precedence rules:

  • If set, it takes precedence over the deprecated proxy-address, proxy-user, proxy-password and non-proxy-hosts properties, which are ignored.

  • If not set and a default proxy configuration is configured (quarkus.proxy.*), then that one is used.

  • If set, the configuration from quarkus.proxy.<name>.* is used.

  • If set but no proxy configuration is found with that name, an error is thrown at runtime.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_PROXY_CONFIGURATION_NAME

string

If true, the REST clients will not provide additional contextual information (like REST client class and method names) when exception occurs during a client invocation.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_DISABLE_CONTEXTUAL_ERROR_MESSAGES

boolean

false

The time in ms for which a connection remains unused in the connection pool before being evicted and closed. A timeout of 0 means there is no timeout.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_CONNECTION_TTL

int

The size of the connection pool for this client.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_CONNECTION_POOL_SIZE

int

If set to false disables the keep alive completely.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_KEEP_ALIVE_ENABLED

boolean

true

The class name of the host name verifier. The class must have a public no-argument constructor.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_HOSTNAME_VERIFIER

string

Set whether hostname verification is enabled. Default is enabled. This setting should not be disabled in production as it makes the client vulnerable to MITM attacks.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_VERIFY_HOST

boolean

The trust store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_TRUST_STORE

string

The trust store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_TRUST_STORE_PASSWORD

string

The type of the trust store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_TRUST_STORE_TYPE

string

The key store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_KEY_STORE

string

The key store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_KEY_STORE_PASSWORD

string

The type of the key store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_KEY_STORE_TYPE

string

The name of the TLS configuration to use.

If not set and the default TLS configuration is configured (quarkus.tls.*) then that will be used. If a name is configured, it uses the configuration from quarkus.tls.<name>.* If a name is configured, but no TLS configuration is found with that name then an error will be thrown.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_TLS_CONFIGURATION_NAME

string

The unique identifier of the guardrail that you want to use.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_GUARDRAIL_GUARDRAIL_IDENTIFIER

string

The version number for the guardrail.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_GUARDRAIL_GUARDRAIL_VERSION

string

The processing mode for streaming requests. Possible values are SYNC or ASYNC.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_GUARDRAIL_STREAM_PROCESSING_MODE

sync, async

Enables prompt caching to reduce latency and costs for requests with repeated content. You can specify where to place the cache point in the prompt, possible values are AFTER_SYSTEM, AFTER_USER_MESSAGE or AFTER_TOOLS.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_PROMPT_CACHING

after-system, after-user-message, after-last-user-message, after-tools

How long the prompt cache entries live. Only applies when prompt-caching is set.

Allowable values: [5m, 1h] (the SDK constant names VALUE_5_M and VALUE_1_H are accepted as well)

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_PROMPT_CACHING_TTL

VALUE_5_M, VALUE_1_H, UNKNOWN_TO_SDK_VERSION

5m

Enables reasoning capabilities of the model. It requires to set the maximum number of tokens to allocate for reasoning.

This uses the legacy budget-based reasoning configuration (reasoning_config.type=enabled with budget_tokens), which is not accepted by newer models such as Claude Opus 4.7+. For those models use reasoning-effort instead.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_REASONING

int

Enables adaptive reasoning with the given effort level (reasoning_config.type=adaptive with output_config.effort). This is the only reasoning configuration accepted by newer models such as Claude Opus 4.7+, which reject the legacy budget-based reasoning configuration. When set, it takes precedence over the reasoning token budget.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_REASONING_EFFORT

low, medium, high

Whether embedding model requests should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_LOG_REQUESTS

boolean

false

Whether embedding model responses should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_LOG_RESPONSES

boolean

false

Whether embedding model body in request and response should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_LOG_BODY

boolean

false

Region used by the bedrock runtime api. See Regions Supported.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_AWS_REGION

string

default aws region provider chain

Override the endpoint used by the bedrock client

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_AWS_ENDPOINT_OVERRIDE

string

Specify a custom credentials provider to use for the bedrock client. Identified by bean name.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_AWS_CREDENTIALS_PROVIDER

string

default aws credentials provider chain

The maximum number retries the aws sdk client will attempt.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_AWS_MAX_RETRIES

int

3

Configure the amount of time to allow the client to complete the execution of an API call. This timeout covers the entire client execution except for marshalling. This includes request handler execution, all HTTP requests including retries, unmarshalling, etc. This value should always be positive, if present.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_AWS_API_CALL_TIMEOUT

Duration 

13s

Model name to use

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_MODEL_ID

string

cohere.embed-english-v3

The number of dimensions the output embedding should have

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_TITAN_DIMENSIONS

int

Flag indicating whether to normalize the output embedding

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_TITAN_NORMALIZE

boolean

Prepends special tokens to differentiate each type from one another. You should not mix different types together, except when mixing types for search and retrieval. In this case, embed your corpus with the search_document type and embedded queries with type search_query type.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_COHERE_INPUT_TYPE

string

Specifies how the API handles inputs longer than the maximum token length

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_COHERE_TRUNCATE

string

Connect Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_CONNECT_TIMEOUT

Duration 

3s

Read Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_TIMEOUT

Duration 

10s

The name of the proxy configuration to use, as registered in the Quarkus Proxy Registry.

Precedence rules:

  • If set, it takes precedence over the deprecated proxy-address, proxy-user, proxy-password and non-proxy-hosts properties, which are ignored.

  • If not set and a default proxy configuration is configured (quarkus.proxy.*), then that one is used.

  • If set, the configuration from quarkus.proxy.<name>.* is used.

  • If set but no proxy configuration is found with that name, an error is thrown at runtime.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_PROXY_CONFIGURATION_NAME

string

If true, the REST clients will not provide additional contextual information (like REST client class and method names) when exception occurs during a client invocation.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_DISABLE_CONTEXTUAL_ERROR_MESSAGES

boolean

false

The time in ms for which a connection remains unused in the connection pool before being evicted and closed. A timeout of 0 means there is no timeout.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_CONNECTION_TTL

int

The size of the connection pool for this client.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_CONNECTION_POOL_SIZE

int

If set to false disables the keep alive completely.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_KEEP_ALIVE_ENABLED

boolean

true

The class name of the host name verifier. The class must have a public no-argument constructor.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_HOSTNAME_VERIFIER

string

Set whether hostname verification is enabled. Default is enabled. This setting should not be disabled in production as it makes the client vulnerable to MITM attacks.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_VERIFY_HOST

boolean

The trust store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_TRUST_STORE

string

The trust store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_TRUST_STORE_PASSWORD

string

The type of the trust store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_TRUST_STORE_TYPE

string

The key store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_KEY_STORE

string

The key store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_KEY_STORE_PASSWORD

string

The type of the key store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_KEY_STORE_TYPE

string

The name of the TLS configuration to use.

If not set and the default TLS configuration is configured (quarkus.tls.*) then that will be used. If a name is configured, it uses the configuration from quarkus.tls.<name>.* If a name is configured, but no TLS configuration is found with that name then an error will be thrown.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_TLS_CONFIGURATION_NAME

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

A string value in the form of : that specifies the HTTP proxy server hostname (or IP address) and port for requests of clients to use.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_PROXY_ADDRESS

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

Proxy username, equivalent to the http.proxy or https.proxy JVM settings.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_PROXY_USER

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

Proxy password, equivalent to the http.proxyPassword or https.proxyPassword JVM settings.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_PROXY_PASSWORD

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

Hosts to access without proxy, similar to the http.nonProxyHosts or https.nonProxyHosts JVM settings. Please note that unlike the JVM settings, this property is empty by default.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CLIENT_NON_PROXY_HOSTS

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

A string value in the form of : that specifies the HTTP proxy server hostname (or IP address) and port for requests of clients to use.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_PROXY_ADDRESS

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

Proxy username, equivalent to the http.proxy or https.proxy JVM settings.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_PROXY_USER

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

Proxy password, equivalent to the http.proxyPassword or https.proxyPassword JVM settings.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_PROXY_PASSWORD

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

Hosts to access without proxy, similar to the http.nonProxyHosts or https.nonProxyHosts JVM settings. Please note that unlike the JVM settings, this property is empty by default.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_CHAT_MODEL_CLIENT_NON_PROXY_HOSTS

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

A string value in the form of : that specifies the HTTP proxy server hostname (or IP address) and port for requests of clients to use.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_PROXY_ADDRESS

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

Proxy username, equivalent to the http.proxy or https.proxy JVM settings.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_PROXY_USER

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

Proxy password, equivalent to the http.proxyPassword or https.proxyPassword JVM settings.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_PROXY_PASSWORD

string

This property is deprecated: Use proxy-configuration-name with the Quarkus Proxy Registry instead.

Hosts to access without proxy, similar to the http.nonProxyHosts or https.nonProxyHosts JVM settings. Please note that unlike the JVM settings, this property is empty by default.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK_EMBEDDING_MODEL_CLIENT_NON_PROXY_HOSTS

string

Named model config

Type

Default

Whether to enable the integration. Defaults to true, which means requests are made to Bedrock. Set to false to disable all requests.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__ENABLE_INTEGRATION

boolean

true

Whether the Bedrock client should log requests

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__LOG_REQUESTS

boolean

false

Whether the Bedrock client should log responses

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__LOG_RESPONSES

boolean

false

Whether chat model body in request and response should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__LOG_BODY

boolean

false

Region used by the bedrock runtime api. See Regions Supported.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__AWS_REGION

string

default aws region provider chain

Override the endpoint used by the bedrock client

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__AWS_ENDPOINT_OVERRIDE

string

Specify a custom credentials provider to use for the bedrock client. Identified by bean name.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__AWS_CREDENTIALS_PROVIDER

string

default aws credentials provider chain

The maximum number retries the aws sdk client will attempt.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__AWS_MAX_RETRIES

int

3

Configure the amount of time to allow the client to complete the execution of an API call. This timeout covers the entire client execution except for marshalling. This includes request handler execution, all HTTP requests including retries, unmarshalling, etc. This value should always be positive, if present.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__AWS_API_CALL_TIMEOUT

Duration 

13s

Connect Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_CONNECT_TIMEOUT

Duration 

3s

Read Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_TIMEOUT

Duration 

10s

The name of the proxy configuration to use, as registered in the Quarkus Proxy Registry.

Precedence rules:

  • If set, it takes precedence over the deprecated proxy-address, proxy-user, proxy-password and non-proxy-hosts properties, which are ignored.

  • If not set and a default proxy configuration is configured (quarkus.proxy.*), then that one is used.

  • If set, the configuration from quarkus.proxy.<name>.* is used.

  • If set but no proxy configuration is found with that name, an error is thrown at runtime.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_PROXY_CONFIGURATION_NAME

string

If true, the REST clients will not provide additional contextual information (like REST client class and method names) when exception occurs during a client invocation.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_DISABLE_CONTEXTUAL_ERROR_MESSAGES

boolean

false

The time in ms for which a connection remains unused in the connection pool before being evicted and closed. A timeout of 0 means there is no timeout.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_CONNECTION_TTL

int

The size of the connection pool for this client.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_CONNECTION_POOL_SIZE

int

If set to false disables the keep alive completely.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_KEEP_ALIVE_ENABLED

boolean

true

The class name of the host name verifier. The class must have a public no-argument constructor.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_HOSTNAME_VERIFIER

string

Set whether hostname verification is enabled. Default is enabled. This setting should not be disabled in production as it makes the client vulnerable to MITM attacks.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_VERIFY_HOST

boolean

The trust store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_TRUST_STORE

string

The trust store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_TRUST_STORE_PASSWORD

string

The type of the trust store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_TRUST_STORE_TYPE

string

The key store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_KEY_STORE

string

The key store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_KEY_STORE_PASSWORD

string

The type of the key store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_KEY_STORE_TYPE

string

The name of the TLS configuration to use.

If not set and the default TLS configuration is configured (quarkus.tls.*) then that will be used. If a name is configured, it uses the configuration from quarkus.tls.<name>.* If a name is configured, but no TLS configuration is found with that name then an error will be thrown.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CLIENT_TLS_CONFIGURATION_NAME

string

Whether chat model requests should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_LOG_REQUESTS

boolean

false

Whether chat model responses should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_LOG_RESPONSES

boolean

false

Whether chat model body in request and response should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_LOG_BODY

boolean

false

Region used by the bedrock runtime api. See Regions Supported.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_AWS_REGION

string

default aws region provider chain

Override the endpoint used by the bedrock client

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_AWS_ENDPOINT_OVERRIDE

string

Specify a custom credentials provider to use for the bedrock client. Identified by bean name.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_AWS_CREDENTIALS_PROVIDER

string

default aws credentials provider chain

The maximum number retries the aws sdk client will attempt.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_AWS_MAX_RETRIES

int

3

Configure the amount of time to allow the client to complete the execution of an API call. This timeout covers the entire client execution except for marshalling. This includes request handler execution, all HTTP requests including retries, unmarshalling, etc. This value should always be positive, if present.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_AWS_API_CALL_TIMEOUT

Duration 

13s

Model id to use. See Models Supported

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_MODEL_ID

string

global.amazon.nova-2-lite-v1:0

What sampling temperature to use, between 0.0 and 1.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.

It is generally recommended to set this or the top-k property but not both.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_TEMPERATURE

double

0.7

The maximum number of tokens to generate in the completion.

The token count of your prompt plus max_tokens cannot exceed the model’s context length

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_MAX_TOKENS

int

1024

Double (0.0-1.0). Nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

It is generally recommended to set this or the temperature property but not both.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_TOP_P

double

1.0

Reduces the probability of generating nonsense. A higher value (e.g. 100) will give more diverse answers, while a lower value (e.g. 10) will be more conservative

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_TOP_K

int

The custom text sequences that will cause the model to stop generating

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_STOP_SEQUENCES

list of string

Specifies the desired format for the model’s output.

Allowable values: [text, json]

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_RESPONSE_FORMAT

string

Connect Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_CONNECT_TIMEOUT

Duration 

3s

Read Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_TIMEOUT

Duration 

10s

The name of the proxy configuration to use, as registered in the Quarkus Proxy Registry.

Precedence rules:

  • If set, it takes precedence over the deprecated proxy-address, proxy-user, proxy-password and non-proxy-hosts properties, which are ignored.

  • If not set and a default proxy configuration is configured (quarkus.proxy.*), then that one is used.

  • If set, the configuration from quarkus.proxy.<name>.* is used.

  • If set but no proxy configuration is found with that name, an error is thrown at runtime.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_PROXY_CONFIGURATION_NAME

string

If true, the REST clients will not provide additional contextual information (like REST client class and method names) when exception occurs during a client invocation.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_DISABLE_CONTEXTUAL_ERROR_MESSAGES

boolean

false

The time in ms for which a connection remains unused in the connection pool before being evicted and closed. A timeout of 0 means there is no timeout.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_CONNECTION_TTL

int

The size of the connection pool for this client.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_CONNECTION_POOL_SIZE

int

If set to false disables the keep alive completely.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_KEEP_ALIVE_ENABLED

boolean

true

The class name of the host name verifier. The class must have a public no-argument constructor.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_HOSTNAME_VERIFIER

string

Set whether hostname verification is enabled. Default is enabled. This setting should not be disabled in production as it makes the client vulnerable to MITM attacks.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_VERIFY_HOST

boolean

The trust store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_TRUST_STORE

string

The trust store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_TRUST_STORE_PASSWORD

string

The type of the trust store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_TRUST_STORE_TYPE

string

The key store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_KEY_STORE

string

The key store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_KEY_STORE_PASSWORD

string

The type of the key store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_KEY_STORE_TYPE

string

The name of the TLS configuration to use.

If not set and the default TLS configuration is configured (quarkus.tls.*) then that will be used. If a name is configured, it uses the configuration from quarkus.tls.<name>.* If a name is configured, but no TLS configuration is found with that name then an error will be thrown.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_CLIENT_TLS_CONFIGURATION_NAME

string

The unique identifier of the guardrail that you want to use.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_GUARDRAIL_GUARDRAIL_IDENTIFIER

string

The version number for the guardrail.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_GUARDRAIL_GUARDRAIL_VERSION

string

The processing mode for streaming requests. Possible values are SYNC or ASYNC.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_GUARDRAIL_STREAM_PROCESSING_MODE

sync, async

Enables prompt caching to reduce latency and costs for requests with repeated content. You can specify where to place the cache point in the prompt, possible values are AFTER_SYSTEM, AFTER_USER_MESSAGE or AFTER_TOOLS.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_PROMPT_CACHING

after-system, after-user-message, after-last-user-message, after-tools

How long the prompt cache entries live. Only applies when prompt-caching is set.

Allowable values: [5m, 1h] (the SDK constant names VALUE_5_M and VALUE_1_H are accepted as well)

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_PROMPT_CACHING_TTL

VALUE_5_M, VALUE_1_H, UNKNOWN_TO_SDK_VERSION

5m

Enables reasoning capabilities of the model. It requires to set the maximum number of tokens to allocate for reasoning.

This uses the legacy budget-based reasoning configuration (reasoning_config.type=enabled with budget_tokens), which is not accepted by newer models such as Claude Opus 4.7+. For those models use reasoning-effort instead.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_REASONING

int

Enables adaptive reasoning with the given effort level (reasoning_config.type=adaptive with output_config.effort). This is the only reasoning configuration accepted by newer models such as Claude Opus 4.7+, which reject the legacy budget-based reasoning configuration. When set, it takes precedence over the reasoning token budget.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__CHAT_MODEL_REASONING_EFFORT

low, medium, high

Whether embedding model requests should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_LOG_REQUESTS

boolean

false

Whether embedding model responses should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_LOG_RESPONSES

boolean

false

Whether embedding model body in request and response should be logged

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_LOG_BODY

boolean

false

Region used by the bedrock runtime api. See Regions Supported.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_AWS_REGION

string

default aws region provider chain

Override the endpoint used by the bedrock client

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_AWS_ENDPOINT_OVERRIDE

string

Specify a custom credentials provider to use for the bedrock client. Identified by bean name.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_AWS_CREDENTIALS_PROVIDER

string

default aws credentials provider chain

The maximum number retries the aws sdk client will attempt.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_AWS_MAX_RETRIES

int

3

Configure the amount of time to allow the client to complete the execution of an API call. This timeout covers the entire client execution except for marshalling. This includes request handler execution, all HTTP requests including retries, unmarshalling, etc. This value should always be positive, if present.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_AWS_API_CALL_TIMEOUT

Duration 

13s

Model name to use

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_MODEL_ID

string

cohere.embed-english-v3

The number of dimensions the output embedding should have

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_TITAN_DIMENSIONS

int

Flag indicating whether to normalize the output embedding

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_TITAN_NORMALIZE

boolean

Prepends special tokens to differentiate each type from one another. You should not mix different types together, except when mixing types for search and retrieval. In this case, embed your corpus with the search_document type and embedded queries with type search_query type.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_COHERE_INPUT_TYPE

string

Specifies how the API handles inputs longer than the maximum token length

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_COHERE_TRUNCATE

string

Connect Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_CONNECT_TIMEOUT

Duration 

3s

Read Timeout for Bedrock calls

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_TIMEOUT

Duration 

10s

The name of the proxy configuration to use, as registered in the Quarkus Proxy Registry.

Precedence rules:

  • If set, it takes precedence over the deprecated proxy-address, proxy-user, proxy-password and non-proxy-hosts properties, which are ignored.

  • If not set and a default proxy configuration is configured (quarkus.proxy.*), then that one is used.

  • If set, the configuration from quarkus.proxy.<name>.* is used.

  • If set but no proxy configuration is found with that name, an error is thrown at runtime.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_PROXY_CONFIGURATION_NAME

string

If true, the REST clients will not provide additional contextual information (like REST client class and method names) when exception occurs during a client invocation.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_DISABLE_CONTEXTUAL_ERROR_MESSAGES

boolean

false

The time in ms for which a connection remains unused in the connection pool before being evicted and closed. A timeout of 0 means there is no timeout.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_CONNECTION_TTL

int

The size of the connection pool for this client.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_CONNECTION_POOL_SIZE

int

If set to false disables the keep alive completely.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_KEEP_ALIVE_ENABLED

boolean

true

The class name of the host name verifier. The class must have a public no-argument constructor.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_HOSTNAME_VERIFIER

string

Set whether hostname verification is enabled. Default is enabled. This setting should not be disabled in production as it makes the client vulnerable to MITM attacks.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_VERIFY_HOST

boolean

The trust store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_TRUST_STORE

string

The trust store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_TRUST_STORE_PASSWORD

string

The type of the trust store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_TRUST_STORE_TYPE

string

The key store location. Can point to either a classpath resource or a file.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_KEY_STORE

string

The key store password.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_KEY_STORE_PASSWORD

string

The type of the key store. Defaults to "JKS".

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_KEY_STORE_TYPE

string

The name of the TLS configuration to use.

If not set and the default TLS configuration is configured (quarkus.tls.*) then that will be used. If a name is configured, it uses the configuration from quarkus.tls.<name>.* If a name is configured, but no TLS configuration is found with that name then an error will be thrown.

Environment variable: QUARKUS_LANGCHAIN4J_BEDROCK__MODEL_NAME__EMBEDDING_MODEL_CLIENT_TLS_CONFIGURATION_NAME

string

About the Duration format

To write duration values, use the standard java.time.Duration format. See the Duration#parse() Java API documentation for more information.

You can also use a simplified format, starting with a number:

  • If the value is only a number, it represents time in seconds.

  • If the value is a number followed by ms, it represents time in milliseconds.

In other cases, the simplified format is translated to the java.time.Duration format for parsing:

  • If the value is a number followed by h, m, or s, it is prefixed with PT.

  • If the value is a number followed by d, it is prefixed with P.