The Paris-based artificial intelligence startup Mistral AI has launched a new conversational assistant, Le Chat, and released its most advanced language model to date, Mistral Large. The company has also announced a partnership with Microsoft to integrate its models into Azure AI, making Mistral a major alternative to models like Google GPT-4.

Mistral Large boasts top-tier reasoning capabilities and proficiency in English, French, Spanish, German, and Italian. The model is also designed for complex cross-lingual reasoning tasks, including text comprehension, translation, and code generation.

In comparison to other leading language models, Mistral Large ranks second in the Massive Multi-Task Language Understanding (MMLU) benchmark test, only behind GPT-4. The model also performs exceptionally well in other benchmark tests such as HellaSwag, WinoGrande, Arc Challenge, TriviaQA, and TruthfulQA.

Mistral Large excels in inference, knowledge, encoding, mathematics, and cross-lingual understanding tests. With only 32,000 token context windows compared to GPT-4's 128K, it is already sufficient for most documents of average length.

To showcase the capabilities of Mistral Large and other models, Mistral AI has introduced Le Chat Mistral, a multilingual conversational assistant currently in the testing phase. Le Chat Mistral offers an interactive and enjoyable way to explore Mistral AI's technology, providing a teaching method to understand the company's progress. The assistant can utilize Mistral Large, Mistral Small, or Mistral Next (a prototype model designed for simplicity).

For enterprises, Mistral AI has also launched Le Chat Enterprise, which features self-deployment capabilities and robust governance mechanisms. This product aims to enhance team productivity while ensuring a secure and controllable AI interaction environment.

The company has also signed an agreement with Microsoft to offer its models through Microsoft Azure. This enables Mistral Large to be used with Azure AI Studio and Azure Machine Learning, significantly expanding Mistral's commercial distribution.

Mistral's self-hosted enterprise products will also run on the Azure infrastructure for sensitive use cases. This collaboration further enhances the credibility of models beyond GPT-4 as the application of AI capabilities intensifies in the industry.