sendMessage method

Future<(GroqResponse, GroqUsage)> sendMessage(String prompt, { GroqMessageRole role = GroqMessageRole.user, String? username, bool expectJSON = false, })

Sends a new request message to the chat prompt the message content role the message role DO NOT USE assistant, it is reserved for the AI username the username of the message sender (optional) expectJSON whether to expect a JSON response or not. You need to explain the JSON structure in the prompt for this feature Returns a tuple of the response and the resource usage Example:

final (response, usage) = await chat.sendMessage('Explain the concept of a chatbot');
print(response.choices.first.message); //prints the response message
print(usage.totalTokens); //prints the total tokens used in the response

Implementation

Future<(GroqResponse, GroqUsage)> sendMessage(
  String prompt, {
  GroqMessageRole role = GroqMessageRole.user,
  String? username,
  bool expectJSON = false,
}) async {
  final request =
      GroqMessage(content: prompt, role: role, username: username);
  _streamController.add(RequestChatEvent(request));
  _chatItems.add(RequestChatEvent(request));
  final item = GroqConversationItem(_model, request);
  GroqResponse response;
  GroqUsage usage;
  GroqRateLimitInformation rateLimitInfo;
  try {
    (response, usage, rateLimitInfo) = await GroqApi.getNewChatCompletion(
      apiKey: _apiKey,
      prompt: request,
      chat: this,
      expectJSON: expectJSON,
    );
  } catch (e) {
    _streamController.addError(e);
    rethrow;
  }
  _rateLimitInfo = rateLimitInfo;
  item.setResponse(response, usage);
  _chatItems.add(ResponseChatEvent(response, usage));
  _conversationItems.add(item);
  _streamController.add(ResponseChatEvent(response, usage));
  return (response, usage);
}