@n8n/n8n-nodes-langchain/dist/node-definitions/nodes/n8n-nodes-langchain/lmChatAwsBedrock/v12.ts

/**
 * AWS Bedrock Chat Model Node - Version 1.2
 * Language Model AWS Bedrock
 */


export interface LcLmChatAwsBedrockV12Params {
  authentication?: 'iam' | 'assumeRole' | Expression<string>;
/**
 * The model or inference profile which will generate the completion. &lt;a href="https://docs.aws.amazon.com/bedrock/latest/userguide/foundation-models.html"&gt;Learn more&lt;/a&gt;.
 * @builderHint Prefer the newest Claude Sonnet model (claude-sonnet-4-6 family). The newest models only work through their inference profile ID, which starts with a region prefix (e.g. "eu.anthropic.claude-sonnet-4-6..."). Avoid claude-sonnet-4-5, claude-3.x, and older non-Claude models unless the user asks for them.
 * @loadOptionsMethod listModels
 */
    model?: string | Expression<string>;
/**
 * Additional options to add
 * @default {}
 */
    options?: {
    /** The maximum number of tokens to generate in the completion
     * @default 2000
     */
    maxTokensToSample?: number | Expression<number>;
    /** Controls randomness: Lowering results in less random completions. As the temperature approaches zero, the model will become deterministic and repetitive.
     * @default 0.7
     */
    temperature?: number | Expression<number>;
    /** Controls diversity via nucleus sampling: 0.5 means half of all likelihood-weighted options are considered. We generally recommend altering this or temperature but not both.
     * @default 1
     */
    topP?: number | Expression<number>;
    /** Maximum number of retries to attempt when a request fails
     * @default 2
     */
    maxRetries?: number | Expression<number>;
    /** Maximum amount of time a request is allowed to take in milliseconds. Increase this for long generations; set to 0 to disable.
     * @default 60000
     */
    timeout?: number | Expression<number>;
    /** Model-family-specific inference parameters passed through as JSON (e.g. Claude &lt;code&gt;top_k&lt;/code&gt;/&lt;code&gt;thinking&lt;/code&gt;, Nova &lt;code&gt;inferenceConfig&lt;/code&gt;/&lt;code&gt;reasoningConfig&lt;/code&gt;, Cohere penalties). See the &lt;a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters.html"&gt;AWS model parameters docs&lt;/a&gt;.
     * @default {}
     */
    additionalModelRequestFields?: IDataObject | string | Expression<string>;
    /** Latency optimization mode for the request. "Optimized" can reduce response time for supported models and regions.
     * @default standard
     */
    latency?: 'standard' | 'optimized' | Expression<string>;
    /** Apply an Amazon Bedrock guardrail to requests
     * @default {}
     */
    guardrail?: {
        /** Guardrail
     */
    values?: {
      /** The identifier (ID or ARN) of the guardrail to apply
       */
      guardrailIdentifier?: string | Expression<string>;
      /** The version of the guardrail to apply, e.g. "1". Defaults to the working draft ("DRAFT").
       * @default DRAFT
       */
      guardrailVersion?: string | Expression<string>;
      /** The trace behavior for the guardrail
       * @default disabled
       */
      trace?: 'disabled' | 'enabled' | 'enabled_full' | Expression<string>;
    };
  };
  };
}

export interface LcLmChatAwsBedrockV12Credentials {
  aws: CredentialReference;
  awsAssumeRole: CredentialReference;
}

interface LcLmChatAwsBedrockV12NodeBase {
  type: '@n8n/n8n-nodes-langchain.lmChatAwsBedrock';
  version: 1.2;
}

export type LcLmChatAwsBedrockV12ParamsNode = LcLmChatAwsBedrockV12NodeBase & {
  config: NodeConfig<LcLmChatAwsBedrockV12Params> & { credentials?: LcLmChatAwsBedrockV12Credentials };
};

export type LcLmChatAwsBedrockV12Node = LcLmChatAwsBedrockV12ParamsNode;