Class BedrockFoundationModel

java.lang.Object
software.amazon.jsii.JsiiObject
software.amazon.awscdk.services.bedrock.alpha.BedrockFoundationModel
All Implemented Interfaces:
IBedrockInvokable, software.amazon.jsii.JsiiSerializable

@Generated(value="jsii-pacmak/1.112.0 (build de1bc80)", date="2025-07-24T11:33:25.531Z") @Stability(Experimental) public class BedrockFoundationModel extends software.amazon.jsii.JsiiObject implements IBedrockInvokable
(experimental) Bedrock models.

If you need to use a model name that doesn't exist as a static member, you can instantiate a BedrockFoundationModel object, e.g: new BedrockFoundationModel('my-model').

Example:

 // Create a specialized agent
 Agent customerSupportAgent = Agent.Builder.create(this, "CustomerSupportAgent")
         .instruction("You specialize in answering customer support questions.")
         .foundationModel(BedrockFoundationModel.AMAZON_NOVA_LITE_V1)
         .build();
 // Create an agent alias
 AgentAlias customerSupportAlias = AgentAlias.Builder.create(this, "CustomerSupportAlias")
         .agent(customerSupportAgent)
         .agentAliasName("production")
         .build();
 // Create a main agent that collaborates with the specialized agent
 Agent mainAgent = Agent.Builder.create(this, "MainAgent")
         .instruction("You route specialized questions to other agents.")
         .foundationModel(BedrockFoundationModel.AMAZON_NOVA_LITE_V1)
         .agentCollaboration(Map.of(
                 "type", AgentCollaboratorType.SUPERVISOR,
                 "collaborators", List.of(
                     AgentCollaborator.Builder.create()
                             .agentAlias(customerSupportAlias)
                             .collaborationInstruction("Route customer support questions to this agent.")
                             .collaboratorName("CustomerSupport")
                             .relayConversationHistory(true)
                             .build())))
         .build();
 
  • Field Details

    • AI21_JAMBA_1_5_LARGE_V1

      @Stability(Experimental) public static final BedrockFoundationModel AI21_JAMBA_1_5_LARGE_V1
      (experimental) AI21's Jamba 1.5 Large model optimized for text generation tasks. Suitable for complex language understanding and generation tasks.

      Features:

      • Supports Bedrock Agents integration
      • Optimized for natural language processing
      • Best for: Content generation, summarization, and complex text analysis
    • AI21_JAMBA_1_5_MINI_V1

      @Stability(Experimental) public static final BedrockFoundationModel AI21_JAMBA_1_5_MINI_V1
      (experimental) AI21's Jamba 1.5 Mini model, a lighter version optimized for faster processing. Balances performance with efficiency for general text tasks.

      Features:

      • Supports Bedrock Agents integration
      • Faster response times compared to larger models
      • Best for: Quick text processing, basic content generation
    • AI21_JAMBA_INSTRUCT_V1

      @Stability(Experimental) public static final BedrockFoundationModel AI21_JAMBA_INSTRUCT_V1
      (experimental) AI21's Jamba Instruct model, specifically designed for instruction-following tasks. Optimized for understanding and executing specific instructions.

      Features:

      • Supports Bedrock Agents integration
      • Enhanced instruction understanding
      • Best for: Task-specific instructions, command processing
    • AMAZON_NOVA_LITE_V1

      @Stability(Experimental) public static final BedrockFoundationModel AMAZON_NOVA_LITE_V1
      (experimental) Amazon's Nova Lite model, balancing performance with efficiency.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Optimized for agents
      • Best for: General-purpose language tasks, moderate complexity
    • AMAZON_NOVA_MICRO_V1

      @Stability(Experimental) public static final BedrockFoundationModel AMAZON_NOVA_MICRO_V1
      (experimental) Amazon's Nova Micro model, a lightweight model optimized for efficiency.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Optimized for agents
      • Best for: Quick processing tasks, basic language understanding
    • AMAZON_NOVA_PREMIER_V1

      @Stability(Experimental) public static final BedrockFoundationModel AMAZON_NOVA_PREMIER_V1
      (experimental) Amazon's Nova Premier model, the most advanced in the Nova series.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Optimized for agents
      • Best for: High-end applications, complex analysis, premium performance
    • AMAZON_NOVA_PRO_V1

      @Stability(Experimental) public static final BedrockFoundationModel AMAZON_NOVA_PRO_V1
      (experimental) Amazon's Nova Pro model, offering advanced capabilities for complex tasks.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Optimized for agents
      • Best for: Complex language tasks, professional applications
    • AMAZON_TITAN_PREMIER_V1_0

      @Stability(Experimental) public static final BedrockFoundationModel AMAZON_TITAN_PREMIER_V1_0
      (experimental) Amazon's Titan Text Premier model, designed for high-quality text generation. Offers enhanced capabilities for complex language tasks.

      Features:

      • Supports Bedrock Agents integration
      • Advanced language understanding
      • Best for: Complex content generation, detailed analysis
    • AMAZON_TITAN_TEXT_EXPRESS_V1

      @Stability(Experimental) public static final BedrockFoundationModel AMAZON_TITAN_TEXT_EXPRESS_V1
      (experimental) Amazon's Titan Text Express model, optimized for fast text generation. Provides quick responses while maintaining good quality output.

      Features:

      • Supports Bedrock Agents integration
      • Fast response times
      • Best for: Real-time applications, chatbots, quick content generation
    • ANTHROPIC_CLAUDE_3_5_HAIKU_V1_0

      @Stability(Experimental) public static final BedrockFoundationModel ANTHROPIC_CLAUDE_3_5_HAIKU_V1_0
      (experimental) Anthropic's Claude 3.5 Haiku model, optimized for quick responses. Lightweight model focused on speed and efficiency.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Optimized for agents
      • Best for: Fast responses, lightweight processing
    • ANTHROPIC_CLAUDE_3_5_SONNET_V1_0

      @Stability(Experimental) public static final BedrockFoundationModel ANTHROPIC_CLAUDE_3_5_SONNET_V1_0
      (experimental) Anthropic's Claude 3.5 Sonnet V1 model, balanced performance model. Offers good balance between performance and efficiency.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Optimized for agents
      • Best for: General language tasks, balanced performance
    • ANTHROPIC_CLAUDE_3_5_SONNET_V2_0

      @Stability(Experimental) public static final BedrockFoundationModel ANTHROPIC_CLAUDE_3_5_SONNET_V2_0
      (experimental) Anthropic's Claude 3.5 Sonnet V2 model, optimized for agent interactions. Enhanced version with improved performance and capabilities.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Optimized for agents
      • Best for: Agent-based applications, complex dialogue
    • ANTHROPIC_CLAUDE_3_7_SONNET_V1_0

      @Stability(Experimental) public static final BedrockFoundationModel ANTHROPIC_CLAUDE_3_7_SONNET_V1_0
      (experimental) Anthropic's Claude 3.7 Sonnet model, latest in the Claude 3 series. Advanced language model with enhanced capabilities.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: Complex reasoning, analysis, and content generation
    • ANTHROPIC_CLAUDE_HAIKU_V1_0

      @Stability(Experimental) public static final BedrockFoundationModel ANTHROPIC_CLAUDE_HAIKU_V1_0
      (experimental) Anthropic's Claude Haiku model, optimized for efficiency. Fast and efficient model for lightweight tasks.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Optimized for agents
      • Best for: Quick responses, simple tasks
    • ANTHROPIC_CLAUDE_INSTANT_V1_2

      @Stability(Experimental) public static final BedrockFoundationModel ANTHROPIC_CLAUDE_INSTANT_V1_2
      (experimental) Anthropic's Claude Instant V1.2 model, legacy version. Fast and efficient model optimized for quick responses.

      Features:

      • Supports Bedrock Agents integration
      • Legacy model with EOL date
      • Optimized for agents
      • Best for: Quick responses, simple tasks, legacy applications
    • ANTHROPIC_CLAUDE_OPUS_V1_0

      @Stability(Experimental) public static final BedrockFoundationModel ANTHROPIC_CLAUDE_OPUS_V1_0
      (experimental) Anthropic's Claude Opus model, designed for advanced tasks. High-performance model with extensive capabilities.

      Features:

      • Supports Bedrock Agents integration
      • Optimized for agents
      • Best for: Complex reasoning, research, and analysis
    • ANTHROPIC_CLAUDE_SONNET_V1_0

      @Stability(Experimental) public static final BedrockFoundationModel ANTHROPIC_CLAUDE_SONNET_V1_0
      (experimental) Anthropic's Claude Sonnet model, legacy version. Balanced model for general-purpose tasks.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Legacy model with EOL date
      • Best for: General language tasks, standard applications
    • ANTHROPIC_CLAUDE_V2

      @Stability(Experimental) public static final BedrockFoundationModel ANTHROPIC_CLAUDE_V2
      (experimental) Anthropic's Claude V2 model, legacy version. Original Claude V2 model with broad capabilities.

      Features:

      • Supports Bedrock Agents integration
      • Legacy model with EOL date
      • Optimized for agents
      • Best for: General language tasks, legacy applications
    • ANTHROPIC_CLAUDE_V2_1

      @Stability(Experimental) public static final BedrockFoundationModel ANTHROPIC_CLAUDE_V2_1
      (experimental) Anthropic's Claude V2.1 model, legacy version. Improved version of Claude V2 with enhanced capabilities.

      Features:

      • Supports Bedrock Agents integration
      • Legacy model with EOL date
      • Optimized for agents
      • Best for: General language tasks, legacy applications
    • COHERE_EMBED_ENGLISH_V3

      @Stability(Experimental) public static final BedrockFoundationModel COHERE_EMBED_ENGLISH_V3
      (experimental) Cohere's English embedding model, optimized for English text embeddings. Specialized for semantic understanding of English content.

      Features:

      • Supports Knowledge Base integration
      • 1024-dimensional vectors
      • Supports both floating-point and binary vectors
      • Best for: English text embeddings, semantic search, content similarity
    • COHERE_EMBED_MULTILINGUAL_V3

      @Stability(Experimental) public static final BedrockFoundationModel COHERE_EMBED_MULTILINGUAL_V3
      (experimental) Cohere's Multilingual embedding model, supporting multiple languages. Enables semantic understanding across different languages.

      Features:

      • Supports Knowledge Base integration
      • 1024-dimensional vectors
      • Supports both floating-point and binary vectors
      • Best for: Cross-lingual embeddings, multilingual semantic search
    • DEEPSEEK_R1_V1

      @Stability(Experimental) public static final BedrockFoundationModel DEEPSEEK_R1_V1
      (experimental) Deepseek's R1 model, designed for general language understanding. Balanced model for various language tasks.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: General language tasks, content generation
    • META_LLAMA_3_1_70_B_INSTRUCT_V1

      @Stability(Experimental) public static final BedrockFoundationModel META_LLAMA_3_1_70_B_INSTRUCT_V1
      (experimental) Meta's Llama 3 70B Instruct model, large-scale instruction model. High-capacity model for complex language understanding.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: Complex instructions, advanced language tasks
    • META_LLAMA_3_1_8_B_INSTRUCT_V1

      @Stability(Experimental) public static final BedrockFoundationModel META_LLAMA_3_1_8_B_INSTRUCT_V1
      (experimental) Meta's Llama 3 1.8B Instruct model, compact instruction-following model. Efficient model optimized for instruction-based tasks.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: Lightweight instruction processing, quick responses
    • META_LLAMA_3_2_11_B_INSTRUCT_V1

      @Stability(Experimental) public static final BedrockFoundationModel META_LLAMA_3_2_11_B_INSTRUCT_V1
      (experimental) Meta's Llama 3.2 11B Instruct model, mid-sized instruction model. Balanced model for general instruction processing.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: General instruction tasks, balanced performance
    • META_LLAMA_3_2_1_B_INSTRUCT_V1

      @Stability(Experimental) public static final BedrockFoundationModel META_LLAMA_3_2_1_B_INSTRUCT_V1
      (experimental) Meta's Llama 3.2 1B Instruct model, ultra-lightweight model. Most compact model in the Llama 3.2 series.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: Simple instructions, fastest response times
    • META_LLAMA_3_2_3_B_INSTRUCT_V1

      @Stability(Experimental) public static final BedrockFoundationModel META_LLAMA_3_2_3_B_INSTRUCT_V1
      (experimental) Meta's Llama 3.2 3B Instruct model, compact efficient model. Lightweight model for basic instruction processing.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: Basic instructions, efficient processing
    • META_LLAMA_3_3_70_B_INSTRUCT_V1

      @Stability(Experimental) public static final BedrockFoundationModel META_LLAMA_3_3_70_B_INSTRUCT_V1
      (experimental) Meta's Llama 3.3 70B Instruct model, latest large-scale model. Advanced model with enhanced capabilities.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: Complex reasoning, advanced language tasks
    • META_LLAMA_4_MAVERICK_17_B_INSTRUCT_V1

      @Stability(Experimental) public static final BedrockFoundationModel META_LLAMA_4_MAVERICK_17_B_INSTRUCT_V1
      (experimental) Meta's Llama 4 Maverick 17B Instruct model, innovative mid-sized model. Specialized for creative and dynamic responses.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: Creative tasks, innovative solutions
    • META_LLAMA_4_SCOUT_17_B_INSTRUCT_V1

      @Stability(Experimental) public static final BedrockFoundationModel META_LLAMA_4_SCOUT_17_B_INSTRUCT_V1
      (experimental) Meta's Llama 4 Scout 17B Instruct model, analytical mid-sized model. Focused on precise and analytical responses.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: Analytical tasks, precise responses
    • MISTRAL_7_B_INSTRUCT_V0

      @Stability(Experimental) public static final BedrockFoundationModel MISTRAL_7_B_INSTRUCT_V0
      (experimental) Mistral's 7B Instruct model, efficient instruction-following model. Balanced performance for instruction processing.

      Features:

      • Supports Bedrock Agents integration
      • Optimized for instruction tasks
      • Best for: General instruction processing, balanced performance
    • MISTRAL_LARGE_2402_V1

      @Stability(Experimental) public static final BedrockFoundationModel MISTRAL_LARGE_2402_V1
      (experimental) Mistral's Large 2402 model, high-capacity language model. Advanced model for complex language understanding.

      Features:

      • Supports Bedrock Agents integration
      • Enhanced language capabilities
      • Best for: Complex reasoning, detailed analysis
    • MISTRAL_LARGE_2407_V1

      @Stability(Experimental) public static final BedrockFoundationModel MISTRAL_LARGE_2407_V1
      (experimental) Mistral's Large 2407 model, updated large-scale model. Enhanced version with improved capabilities.

      Features:

      • Supports Bedrock Agents integration
      • Advanced language processing
      • Best for: Sophisticated language tasks, complex analysis
    • MISTRAL_MIXTRAL_8_X7_B_INSTRUCT_V0

      @Stability(Experimental) public static final BedrockFoundationModel MISTRAL_MIXTRAL_8_X7_B_INSTRUCT_V0
      (experimental) Mistral's Mixtral 8x7B Instruct model, mixture-of-experts architecture. Advanced model combining multiple expert networks.

      Features:

      • Supports Bedrock Agents integration
      • Specialized expert networks
      • Best for: Complex tasks, diverse language understanding
    • MISTRAL_PIXTRAL_LARGE_2502_V1

      @Stability(Experimental) public static final BedrockFoundationModel MISTRAL_PIXTRAL_LARGE_2502_V1
      (experimental) Mistral's Pixtral Large 2502 model, specialized large model. Advanced model with cross-region support.

      Features:

      • Supports Bedrock Agents integration
      • Cross-region support
      • Best for: Advanced language tasks, distributed applications
    • MISTRAL_SMALL_2402_V1

      @Stability(Experimental) public static final BedrockFoundationModel MISTRAL_SMALL_2402_V1
      (experimental) Mistral's Small 2402 model, compact efficient model. Optimized for quick responses and efficiency.

      Features:

      • Supports Bedrock Agents integration
      • Efficient processing
      • Best for: Quick responses, basic language tasks
    • TITAN_EMBED_TEXT_V1

      @Stability(Experimental) public static final BedrockFoundationModel TITAN_EMBED_TEXT_V1
      (experimental) Amazon's Titan Embed Text V1 model for text embeddings.

      Features:

      • Supports Knowledge Base integration
      • 1536-dimensional vectors
      • Floating-point vector type
      • Best for: Text embeddings, semantic search, document similarity
    • TITAN_EMBED_TEXT_V2_1024

      @Stability(Experimental) public static final BedrockFoundationModel TITAN_EMBED_TEXT_V2_1024
      (experimental) Amazon's Titan Embed Text V2 model with 1024-dimensional vectors.

      Features:

      • Supports Knowledge Base integration
      • 1024-dimensional vectors
      • Supports both floating-point and binary vectors
      • Best for: High-dimensional embeddings, advanced semantic search
    • TITAN_EMBED_TEXT_V2_256

      @Stability(Experimental) public static final BedrockFoundationModel TITAN_EMBED_TEXT_V2_256
      (experimental) Amazon's Titan Embed Text V2 model with 256-dimensional vectors.

      Features:

      • Supports Knowledge Base integration
      • 256-dimensional vectors
      • Supports both floating-point and binary vectors
      • Best for: Efficient embeddings with lower dimensionality
    • TITAN_EMBED_TEXT_V2_512

      @Stability(Experimental) public static final BedrockFoundationModel TITAN_EMBED_TEXT_V2_512
      (experimental) Amazon's Titan Embed Text V2 model with 512-dimensional vectors.

      Features:

      • Supports Knowledge Base integration
      • 512-dimensional vectors
      • Supports both floating-point and binary vectors
      • Best for: Balanced performance and dimensionality
  • Constructor Details

    • BedrockFoundationModel

      protected BedrockFoundationModel(software.amazon.jsii.JsiiObjectRef objRef)
    • BedrockFoundationModel

      protected BedrockFoundationModel(software.amazon.jsii.JsiiObject.InitializationMode initializationMode)
    • BedrockFoundationModel

      @Stability(Experimental) public BedrockFoundationModel(@NotNull String value, @Nullable BedrockFoundationModelProps props)
      Parameters:
      value - This parameter is required.
      props -
    • BedrockFoundationModel

      @Stability(Experimental) public BedrockFoundationModel(@NotNull String value)
      Parameters:
      value - This parameter is required.
  • Method Details

    • fromCdkFoundationModel

      @Stability(Experimental) @NotNull public static BedrockFoundationModel fromCdkFoundationModel(@NotNull FoundationModel modelId, @Nullable BedrockFoundationModelProps props)
      (experimental) Creates a BedrockFoundationModel from a FoundationModel.

      Use this method when you have a foundation model from the CDK.

      Parameters:
      modelId -
      • The foundation model.
      This parameter is required.
      props -
      • Optional properties for the model.
      Returns:
      A new BedrockFoundationModel instance
    • fromCdkFoundationModel

      @Stability(Experimental) @NotNull public static BedrockFoundationModel fromCdkFoundationModel(@NotNull FoundationModel modelId)
      (experimental) Creates a BedrockFoundationModel from a FoundationModel.

      Use this method when you have a foundation model from the CDK.

      Parameters:
      modelId -
      • The foundation model.
      This parameter is required.
      Returns:
      A new BedrockFoundationModel instance
    • fromCdkFoundationModelId

      @Stability(Experimental) @NotNull public static BedrockFoundationModel fromCdkFoundationModelId(@NotNull FoundationModelIdentifier modelId, @Nullable BedrockFoundationModelProps props)
      (experimental) Creates a BedrockFoundationModel from a FoundationModelIdentifier.

      Use this method when you have a model identifier from the CDK.

      Parameters:
      modelId -
      • The foundation model identifier.
      This parameter is required.
      props -
      • Optional properties for the model.
      Returns:
      A new BedrockFoundationModel instance
    • fromCdkFoundationModelId

      @Stability(Experimental) @NotNull public static BedrockFoundationModel fromCdkFoundationModelId(@NotNull FoundationModelIdentifier modelId)
      (experimental) Creates a BedrockFoundationModel from a FoundationModelIdentifier.

      Use this method when you have a model identifier from the CDK.

      Parameters:
      modelId -
      • The foundation model identifier.
      This parameter is required.
      Returns:
      A new BedrockFoundationModel instance
    • asArn

      @Stability(Experimental) @NotNull public String asArn()
      (experimental) Returns the ARN of the foundation model in the following format: arn:${Partition}:bedrock:${Region}::foundation-model/${ResourceId}.
    • asIModel

      @Stability(Experimental) @NotNull public IModel asIModel()
      (experimental) Returns the IModel.
    • grantInvoke

      @Stability(Experimental) @NotNull public Grant grantInvoke(@NotNull IGrantable grantee)
      (experimental) Gives the appropriate policies to invoke and use the Foundation Model in the stack region.

      Specified by:
      grantInvoke in interface IBedrockInvokable
      Parameters:
      grantee - This parameter is required.
    • grantInvokeAllRegions

      @Stability(Experimental) @NotNull public Grant grantInvokeAllRegions(@NotNull IGrantable grantee)
      (experimental) Gives the appropriate policies to invoke and use the Foundation Model in all regions.

      Parameters:
      grantee - This parameter is required.
    • toString

      @Stability(Experimental) @NotNull public String toString()
      (experimental) Returns a string representation of an object.
      Overrides:
      toString in class Object
    • getInvokableArn

      @Stability(Experimental) @NotNull public String getInvokableArn()
      (experimental) The ARN used for invoking the model.

      This is the same as modelArn for foundation models.

      Specified by:
      getInvokableArn in interface IBedrockInvokable
    • getModelArn

      @Stability(Experimental) @NotNull public String getModelArn()
      (experimental) The ARN of the foundation model.

      Format: arn:${Partition}:bedrock:${Region}::foundation-model/${ResourceId}

    • getModelId

      @Stability(Experimental) @NotNull public String getModelId()
      (experimental) The unique identifier of the foundation model.
    • getSupportsAgents

      @Stability(Experimental) @NotNull public Boolean getSupportsAgents()
      (experimental) Whether this model supports integration with Bedrock Agents.

      When true, the model can be used with Bedrock Agents for automated task execution.

    • getSupportsCrossRegion

      @Stability(Experimental) @NotNull public Boolean getSupportsCrossRegion()
      (experimental) Whether this model supports cross-region inference.

      When true, the model can be used with Cross-Region Inference Profiles.

    • getSupportsKnowledgeBase

      @Stability(Experimental) @NotNull public Boolean getSupportsKnowledgeBase()
      (experimental) Whether this model supports integration with Bedrock Knowledge Base.

      When true, the model can be used for knowledge base operations.

    • getSupportedVectorType

      @Stability(Experimental) @Nullable public List<VectorType> getSupportedVectorType()
      (experimental) The vector types supported by this model for embeddings.

      Defines whether the model supports floating-point or binary vectors.

    • getVectorDimensions

      @Stability(Experimental) @Nullable public Number getVectorDimensions()
      (experimental) The dimensionality of the vector embeddings produced by this model.

      Only applicable for embedding models.