Mistral AI, a French startup, has officially closed its much-anticipated Series A funding round. The company has raised €385 million, or $415 million at today’s exchange rate, valuing the company at around $2 billion, according to Bloomberg. Mistral AI is also making its commercial platform available today.
Mistral AI, co-founded by Google’s DeepMind and Meta alums, raised a $112 million seed round less than six months ago to establish a European rival to OpenAI. Mistral AI is working on foundational models with an open technology angle.
Andreessen Horowitz (a16z) is leading the latest funding round, with Lightspeed Venture Partners investing in the AI company once more. Salesforce, BNP Paribas, CMA-CGM, General Catalyst, Elad Gil, and Conviction are among the many investors taking part in the round.
“Since the creation of Mistral AI in May, we have been pursuing a clear trajectory: that of creating a European champion with a global vocation in generative artificial intelligence, based on an open, responsible and decentralised approach to technology,” Arthur Mensch, co-founder and CEO of Mistral AI, said in a statement released on Monday.
Mistral AI debuted its first model, the Mistral 7B, in September. Because it was trained on a “small” dataset of around 7 billion parameters, this large language model isn’t meant to compete directly with GPT-4 or Claude 2.
Rather than providing API access to the Mistral 7B model, the company made it available as a free download for developers to run on their devices and servers.
The model was distributed under the Apache 2.0 license, which is an open-source license with no restrictions on use or reproduction other than attribution. While the model can be run by anyone, it was created behind closed doors using a proprietary dataset and weights that are not disclosed.
Mistral AI was also instrumental in shaping the discussions surrounding the EU’s AI Act. The French AI startup has been lobbying for a complete exemption for foundational models, arguing that regulation should only apply to use cases and companies working on products that are directly used by end users.
Just a few days ago, EU lawmakers reached a political agreement. Companies developing foundational models will be subject to transparency requirements, including the sharing of technical documentation and summaries of what’s in the datasets.
Mistral AI’s best model is now only available through an API.
The company intends to continue making money from its foundational models. As a result, Mistral AI is launching its developer platform in beta today. Other companies will be able to pay to use Mistral AI’s models via APIs through this platform.
Developers will be able to access the new Mixtral 8x7B model (“Mistral-small”) in addition to the Mistral 7B model (“Mistral-tiny”). This model employs “a router network” to process input tokens and select the most appropriate set of parameters to provide an answer.
“This technique increases the number of parameters of a model while controlling cost and latency, as the model only uses a fraction of the total set of parameters per token. Concretely, Mixtral has 45B total parameters but only uses 12B parameters per token. It, therefore, processes input and generates output at the same speed and for the same cost as a 12B model,” the company wrote in a blog post.
Mixtral 8x7B is also released under the Apache 2.0 license and is free to download. Mistral’s developer platform now includes a third model, Mistral-medium. It is said to outperform Mistral AI’s other models, and it is only available through the paid API platform — there is no download link.
