France-based Mistral AI recently revealed its first generative artificial intelligence (AI) model in order to compete with US rivals in AI (which includes Meta Platforms Inc. and Microsoft Corportation-supported OpenAI).
The Paris-headquartered firm is introducing a small-scale language model with around 7 billion parameters which should become available free-of-cost to developers, the firm stated in an update this past Wednesday.
In statement shared with Bloomberg, Mistral AI Chief Executive Officer Arthur Mensch, said
“We have training methods that make us more efficient and two times less expensive to implement.”
As noted in the update, large language models (LLMs) are the core infrastructure for generative AI, a tech that develops content from videos to poetry out of various user commands.
Four French firms are reportedly testing out the model for apps such as customer service, chatbots, summaries as well as marketing content generation. This, according to insights shared by Mensch.
The open-source model, known as Mistral 7B, may work well with English as well as computer source code, and may be followed up by more advanced models and commercial offers, he explained.
Since the launch of OpenAI’s ChatGPT this past year, investors have been allocating funds into the potentially disruptive tech. Notably, Microsoft has now invested around $13 billion into OpenAI.
Mistral, which employs 18 workers, was established in 2023 by several researchers including: Timothée Lacroix and Guillaume Lample — who had worked at Meta prior to launching this initiative — and Mensch, who previously worked at Google’s DeepMind.
The firm secured €105 million (appr. $110 million) in a funding round that was finalized in June 2023 (one of the biggest seed rounds ever for a European generative AI firm).
During the past year, generative AI solutions have been adopted across several key sectors including Fintech and other financial services platforms. This new innovation aims to improve the efficiency of products and services, however, the tech is still in its early stages of development and it may take more time to fully understand its true potential.
