Class BaseLLMAbstract

Protected _additionalSettings

Protected thinkingStreamState

Protected addThinkingToMessage

• addThinkingToMessage(message, thinkingContent?): ChatCompletionMessage

• Add thinking content to a chat completion message

  Parameters

  • message: ChatCompletionMessage
  • Optional thinkingContent: string

  Returns ChatCompletionMessage

Protected Abstract createStreamingRequest

• createStreamingRequest(params): Promise<any>

• Create a provider-specific streaming request

  Parameters

  • params: ChatParams

    Chat parameters

  Returns Promise<any>

  A stream object that can be iterated with for await

Protected extractThinkingFromContent

• extractThinkingFromContent(content): {
      content: string;
      thinking?: string;
  }

• Extract thinking content from non-streaming content This method handles case-insensitive extraction of thinking blocks

  Parameters

  • content: string

  Returns {
      content: string;
      thinking?: string;
  }

  • content: string

  • Optional thinking?: string

Protected Abstract finalizeStreamingResponse

• finalizeStreamingResponse(accumulatedContent, lastChunk, usage): ChatResult

• Create the final response object from streaming results

  Parameters

  • accumulatedContent: string

    The complete content accumulated from all chunks
  • lastChunk: any

    The last chunk received from the stream
  • usage: any

    The usage information (tokens, etc.)

  Returns ChatResult

  A complete ChatResult object

Protected getThinkingTagFormat

• getThinkingTagFormat(): {
      close: string;
      open: string;
  }

• Get the thinking tag format for this provider Providers can override this to customize the thinking tag format

  Returns {
      close: string;
      open: string;
  }

  • close: string

  • open: string

Protected handleStreamingChatCompletion

• handleStreamingChatCompletion(params): Promise<ChatResult>

• Template method for handling streaming chat completion This implements the common pattern across providers while delegating provider-specific logic to abstract methods.

  Parameters

  • params: ChatParams

  Returns Promise<ChatResult>

Protected initializeThinkingStreamState

• initializeThinkingStreamState(): void

• Initialize thinking stream state for streaming extraction

  Returns void

Protected Abstract nonStreamingChatCompletion

• nonStreamingChatCompletion(params): Promise<ChatResult>

• Implementation for non-streaming chat completion

  Parameters

  • params: ChatParams

  Returns Promise<ChatResult>

Protected processStreamChunkWithThinking

• processStreamChunkWithThinking(rawContent): string

• Process streaming chunk with thinking extraction This method handles case-insensitive extraction across chunk boundaries

  Parameters

  • rawContent: string

  Returns string

Protected Abstract processStreamingChunk

• processStreamingChunk(chunk): {
      content: string;
      finishReason?: string;
      usage?: any;
  }

• Process a streaming chunk from the provider

  Parameters

  • chunk: any

    The raw chunk from the provider

  Returns {
      content: string;
      finishReason?: string;
      usage?: any;
  }

  Processed content and metadata

  • content: string

  • Optional finishReason?: string

  • Optional usage?: any

Protected resetThinkingStreamState

• resetThinkingStreamState(): void

• Reset thinking stream state

  Returns void

Protected supportsThinkingModels

• supportsThinkingModels(): boolean

• Check if the provider supports thinking models Providers should override this to return true if they support thinking extraction

  Returns boolean