French artificial intelligence startup Mistral AI is set to raise about €450 million ($487 million) in its latest funding round.
The round includes investments from prominent companies such as Nvidia and Salesforce, and is expected to value Mistral AI at around $2 billion, making it a strong competitor to OpenAI, Bloomberg reported Monday (Dec. 4), citing sources familiar with the deal.
The funding round is being led by Andreessen Horowitz, which is reportedly in discussions to invest €200 million ($216.9 million) in funding. Mistral AI will receive over €325 million ($352 million) in equity from investors.
As part of the funding round, Mistral AI’s three co-founders have agreed to sell more than €1 million ($1.08 million) each in equity, per the report.
Mistral AI, founded by former scientists from Google’s DeepMind and Meta, has quickly gained prominence in the European artificial intelligence (AI) startup scene. The company specializes in open-source software for chatbots and generative AI tools, leveraging the experience of its founders in working on large language models similar to those developed by OpenAI.
The startup was launched in May and valued at $259 million in June, PYMTS reported at the time, following a record $113 million in seed funding, the largest-ever seed round recorded in Europe.
Mistral unveiled its first generative AI model in September. Mistral 7B is a “small-sized” model with 7 billion parameters, and will be available for free to developers, PYMNTS reported.
“Mistral 7B is only a first step toward building the frontier models on our roadmap,” Mistral AI said in a press release. “Yet, it can be used to solve many tasks: summarization, structuration and question answering to name a few. It processes and generates text much faster than large proprietary solutions and runs at a fraction of their costs.”
CEO Arthur Mensch said at the time that it had implemented a more efficient and cost-effective training method than competitors such as OpenAI’s ChatGPT, positioning itself as a strong contender in the AI market.
“We have training methods that make us more efficient and two times less expensive to implement,” Mensch said.
SOURCE: PYMNTS