Method SendPromptStream

llm/client/streaming_client.go:58–58

SendPromptStream sends a prompt and returns a channel of streaming chunks. The channel is closed when streaming completes or an error occurs. maxTokens of 0 means use provider default.

(ctx context.Context, prompt string, history []models.Message, maxTokens int)

	// The channel is closed when streaming completes or an error occurs.
	// maxTokens of 0 means use provider default.
	SendPromptStream(ctx context.Context, prompt string, history []models.Message, maxTokens int) (<-chan StreamChunk, error)

	// SupportsStreaming returns true if the provider supports streaming.
	// Some providers may conditionally support streaming (e.g., only for certain models).