Mistral AI unveils Mistral Large, a new LLM that surpasses all but GPT-4

Mistral AI, a company specializing in artificial intelligence, has unveiled two new language models: Mistral Large and Mistral Small. This comes after Microsoft and Mistral announced a new partnership to accelerate

Mistral Large is a high-performance model designed for complex tasks such as multilingual reasoning, text comprehension, and code generation. The model’s capabilities are benchmarked against existing models, demonstrating competitive performance in reasoning, knowledge retrieval, and understanding multiple languages.

Here are some key characteristics of Mistral Large:

• Demonstrated proficiency in reasoning and knowledge tasks.
• Understand and respond in English, French, Spanish, German, and Italian.
• Capacity to retain information from extensive documents.
• Features designed to facilitate developer control and system integration.

Alongside Mistral Large, Mistral AI introduces Mistral Small. This model prioritizes efficiency and cost-effectiveness, making it suitable for applications requiring fast response times and resource optimization.

Both models are available through la Plateforme, a secure platform Mistral offers, and Microsoft Azure. Additionally, organizations with specific security requirements can choose to self-deploy the models. Mistral also announced its AI-powered chatbot, Le Chat. 

Mistral AI has streamlined its endpoint options to cater to diverse performance and cost needs. This update allows developers to select the most suitable option for their projects.

Launching these models and platform enhancements represents Mistral AI’s contribution to the evolving landscape of large language models. The company’s focus on accessibility and performance positions it as a potential player in driving innovation across various industries.

More here.