@n8n/n8n-nodes-langchain

Version:
90 lines (84 loc) • 4.54 kB
text/typescript
/**
 * Anthropic Chat Model Node - Version 1.2
 * Language Model Anthropic
 */


export interface LcLmChatAnthropicV12Params {
/**
 * The model which will generate the completion. &lt;a href="https://docs.anthropic.com/claude/docs/models-overview"&gt;Learn more&lt;/a&gt;.
 * @builderHint This node version only offers superseded Claude models. Pick claude-3-5-sonnet-20241022 if the node has to stay on this version; otherwise rebuild it on the latest node version, where the current Claude generation is selectable.
 * @default claude-3-5-sonnet-20240620
 */
    model?: 'claude-3-5-sonnet-20241022' | 'claude-3-opus-20240229' | 'claude-3-5-sonnet-20240620' | 'claude-3-sonnet-20240229' | 'claude-3-5-haiku-20241022' | 'claude-3-haiku-20240307' | Expression<string>;
/**
 * Additional options to add
 * @default {}
 */
    options?: {
    /** The maximum number of tokens to generate in the completion
     * @default 4096
     */
    maxTokensToSample?: number | Expression<number>;
    /** Controls randomness: Lowering results in less random completions. As the temperature approaches zero, the model will become deterministic and repetitive. Not supported on newer Anthropic models (Claude Opus 4.7+, Claude Sonnet 5+) — ignored there.
     * @displayOptions.hide { thinking: [true], thinkingMode: ["adaptive", "manual"] }
     * @default 0.7
     */
    temperature?: number | Expression<number>;
    /** Used to remove "long tail" low probability responses. Defaults to -1, which disables it. Not supported on newer Anthropic models (Claude Opus 4.7+, Claude Sonnet 5+) — ignored there.
     * @displayOptions.hide { thinking: [true], thinkingMode: ["adaptive", "manual"] }
     * @default -1
     */
    topK?: 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. Not supported on newer Anthropic models (Claude Opus 4.7+, Claude Sonnet 5+) — ignored there.
     * @displayOptions.hide { thinking: [true], thinkingMode: ["adaptive", "manual"] }
     * @default 1
     */
    topP?: number | Expression<number>;
    /** Whether to enable thinking mode for the model
     * @default false
     */
    thinking?: boolean | Expression<boolean>;
    /** The maximum number of tokens to use for thinking
     * @displayOptions.show { thinking: [true] }
     * @default 1024
     */
    thinkingBudget?: number | Expression<number>;
    /** How extended thinking is configured. Leave the option unset to use the model default.
     * @default disabled
     */
    thinkingMode?: 'disabled' | 'adaptive' | 'manual' | Expression<string>;
    /** Effort level for adaptive thinking
     * @displayOptions.show { thinkingMode: ["adaptive"], /model.value: [{"_cnd":{"includes":"opus"}}] }
     * @default medium
     */
    effort?: 'low' | 'medium' | 'high' | 'xhigh' | 'max' | Expression<string>;
    /** Effort level for adaptive thinking
     * @displayOptions.show { thinkingMode: ["adaptive"], /model.value: [{"_cnd":{"regex":"^(?!.*opus).*"}}] }
     * @default medium
     */
    effort?: 'low' | 'medium' | 'high' | Expression<string>;
    /** Maximum tokens used for thinking. Manual mode is rejected by Opus 4.7+.
     * @displayOptions.show { thinkingMode: ["manual"] }
     * @default 1024
     */
    thinkingBudget?: number | Expression<number>;
    /** Whether the model should stream its response over Server-Sent Events instead of returning a single non-streamed payload. Final output shape is unchanged.
     * @default false
     */
    streaming?: boolean | Expression<boolean>;
    /** Whether to cache the system prompt, tool definitions, and conversation history between requests using &lt;a href="https://docs.anthropic.com/en/docs/build-with-claude/prompt-caching"&gt;Anthropic prompt caching&lt;/a&gt;. The value sets how long cached content stays valid before it has to be written again.
     * @default disabled
     */
    promptCaching?: 'disabled' | '5m' | '1h' | Expression<string>;
  };
}

export interface LcLmChatAnthropicV12Credentials {
  anthropicApi: CredentialReference;
}

interface LcLmChatAnthropicV12NodeBase {
  type: '@n8n/n8n-nodes-langchain.lmChatAnthropic';
  version: 1.2;
}

export type LcLmChatAnthropicV12ParamsNode = LcLmChatAnthropicV12NodeBase & {
  config: NodeConfig<LcLmChatAnthropicV12Params> & { credentials?: LcLmChatAnthropicV12Credentials };
};

export type LcLmChatAnthropicV12Node = LcLmChatAnthropicV12ParamsNode;