Mistral AI unveils open-source language model Mistral 7B

The Paris-based startup Mistral AI developed a large open-source language model, Mistral 7B. It is positioned as a more powerful rival to major players such as OpenAI’s ChatGPT.

According to the developers, the new technology outperforms its rivals on nearly every parameter.

Mistral places emphasis on openness and the adaptability of the tool.

«We are releasing Mistral 7B under the Apache 2.0 licence; it can be used without restrictions. Working with open models is the best way for providers and users to build a sustainable business on AI solutions», said the company.

This approach promotes the democratization of AI and transparency, and it also allows other teams to audit, adapt, and improve the existing codebase, according to the developers.

Mistral also provides additional privacy for users, since the neural network can be run on local machines and does not require internet access or certain cloud services.

The LMM code is available in the repository HugginFace.

«Mistral 7B is a demonstration that a language model can be easily tuned to achieve high performance. But it lacks a moderation mechanism. We look forward to working with the community to find ways to ensure the model adheres to constraints for deployment in ecosystems that require moderated outputs», the company noted.

Earlier, Meta introduced a suite of AI tools — the company outlined a new voice assistant, a neural network with various personalities, ‘smart glasses’, and stickers.

In September, Microsoft announced Copilot, a system that unifies Windows OS interfaces with AI language models.