ChatMistralAI

Wrapper around Mistral AI Chat Completions API.

Mistral AI brings the strongest open generative models to the developers, along with efficient ways to deploy and customise them for production.

Note: Mistral AI API is currently in closed beta. You can request access here.

Setup

To use ChatMistralAI you need to have a Mistral AI account and an API key. You can get one here.

The following models are available at the moment:

• mistral-tiny: Mistral 7B Instruct v0.2 (a minor release of Mistral 7B Instruct). It only works in English and obtains 7.6 on MT-Bench.
• mistral-small: Mixtral 8x7B. It masters English/French/Italian/German/Spanish and code and obtains 8.3 on MT-Bench.
• mistral-medium: a prototype model, that is currently among the top serviced models available based on standard benchmarks. It masters English/French/Italian/German/Spanish and code and obtains a score of 8.6 on MT-Bench.

Usage

final chatModel = ChatMistralAI(
  apiKey: 'apiKey',
  defaultOptions: ChatMistralAIOptions(
    model: 'mistral-small',
    temperature: 0,
  ),
);

const template = 'You are a helpful assistant that translates {input_language} to {output_language}.';
final systemMessagePrompt = SystemChatMessagePromptTemplate.fromTemplate(template);
const humanTemplate = '{text}';
final humanMessagePrompt = HumanChatMessagePromptTemplate.fromTemplate(humanTemplate);
final chatPrompt = ChatPromptTemplate.fromPromptMessages(
  [systemMessagePrompt, humanMessagePrompt],
);

final chain = chatPrompt | chatModel | StringOutputParser();

final res = await chain.invoke({
  'input_language': 'English',
  'output_language': 'French',
  'text': 'I love programming.',
});
print(res);
// -> 'J'aime la programmation.'

Streaming

final promptTemplate = ChatPromptTemplate.fromPromptMessages([
  SystemChatMessagePromptTemplate.fromTemplate(
    'You are a helpful assistant that replies only with numbers '
        'in order without any spaces or commas',
  ),
  HumanChatMessagePromptTemplate.fromTemplate(
    'List the numbers from 1 to {max_num}',
  ),
]);
final chat = ChatMistralAI(
  apiKey: 'apiKey',
  defaultOptions: ChatMistralAIOptions(
    model: 'mistral-medium',
    temperature: 0,
  ),
);

final chain = promptTemplate.pipe(chat).pipe(StringOutputParser());

final stream = chain.stream({'max_num': '9'});
await stream.forEach(print);
// 12
// 345
// 67
// 89