AI

France’s Mistral AI blows in with a $113M seed round at a $260M valuation to take on OpenAI

Comment

A photo of palm trees bending in category 1 typhoon winds on Boracay Island in the Philippines
Image Credits: John Seaton Callahan (opens in a new window) / Getty Images

AI is well and truly off to the races: a startup that is only four weeks old has picked up a $113 million round of seed funding to compete against OpenAI in the building, training and application of large language models and generative AI.

Mistral AI, based out of Paris, is co-founded by alums from Google’s DeepMind and Meta and will be focusing on open source solutions and targeting enterprises to create what CEO Arthur Mensch believes is currently the biggest challenge in the field: “To make AI useful.” It plans to release its first models for text-based generative AI in 2024.

Lightspeed Venture Partners is leading this round, with Redpoint, Index Ventures, Xavier Niel, JCDecaux Holding, Rodolphe Saadé and Motier Ventures in France, La Famiglia and Headline in Germany, Exor Ventures in Italy, Sofina in Belgium, and First Minute Capital and LocalGlobe in the UK all also participating.

Mistral AI notes that French investment bank Bpifrance and former Google CEO Eric Schmidt are also shareholders. Sources close to the company confirm that the €105 million in funding ($113M at today’s rates) values Mistral AI €240 million ($260 million). To note, this is the same number that was being rumored about a month ago in the French press when people started chattering about the company.

Guillaume LAMPLE. Arthur MENSCH. Timothee LACROIX. Mistral. Paris. France. 05/2023 © david atlan

Mensch (center), along with his co-founders Timothée Lacroix (CTO – L) and Guillaume Lample (Chief Science Officer – R), are all in their early thirties and have known each other since school, when they were all studying across the field of artificial intelligence.

Mensch was working at DeepMind in Paris, and Lacroix and Lample were at Meta’s Paris AI outpost; and Mensch said it was sometime last year that they got to talking about the direction they could see AI development taking.

“We could see the technology really start to accelerate last year,” he said in an interview today, most likely in reference to the leaps that OpenAI was making with its GPT model, which was a shot in the arm for a lot of people in AI and the wider world of tech.

But while OpenAI has the word “open” in its name, it felt like anything but. Mensch, Lacroix and Lample felt that a proprietary approach was largely shaping up to be the norm, and they saw an opportunity to do things differently. “Open source is a core part of our DNA,” Mensch noted.

It’s very early to talk about what Mistral is doing or will be doing — it’s only around a month old — but from what Mensch said, the plan is to build models using only publicly available data to avoid legal issues that some others have faced over training data, he said; users will be able to contribute their own datasets, too. Models and data sets will be open-sourced, as well.

And while some believe that open source has created a tricky landscape (and minefield) when it comes to areas like application security, “we believe that the benefit of using open source can overcome the misuse potential,” he added. “Open source can prove tactical in security and we believe it will be the case here, too.”

It’s also too soon to know how well its future products will resonate in the market. But what’s interesting is the startup’s singular focus on enterprise, not consumer, customers, and the idea that there is a gap in the market for helping those customers figure out what they need to do, and how they can do it.

“At the moment we have proof that AI is useful in some cases,” Mensch said. “But there are still too many workers in different fields being asked to be creative [with AI], and we need to figure this out for them. We want to give them tools that are easy to use to create their own products.”

It may seem like a big leap to give funding to such a young company without any customers, let alone a product, to its name — especially since the penny is already dropping in some high-profile cases. Neeva (another startup with a Google pedigree) gave up on its consumer AI search play and sold a chunk of its tech to Snowflake; and recently Stability AI has also been in the spotlight, though not in a good way.

But Antoine Moyroud, who led the investment for Lightspeed (which also backed Stability AI, I should point out), said that he believes it’s a leap worth taking.

“For us at Lightspeed, we didn’t work in isolation,” he said in an interview. “We had a chance to see a few strong tech teams looking to raise funding to build LLM across India, the U.S. and Europe. But what was lacking was while they were strong technically they hadn’t thought about the second order effects of LLM.” That is to say, they were not thinking of where and how they might be applied.

He likens the AI landscape to bigger infrastructure plays like cloud computing and database businesses, not applications. “We think value will accrue here as it has in the cloud, where there will be 5-6 players getting the lion’s share.

“Then we thought who are the most credible here? Then we met with the Mistral founders. It’s a very talented team. We think there are only 70-100 people in the world right now with their expertise for language models and optimizing them.” Lample, as one example, led the development of LLaMA, Meta’s Large Language Model, and he was leading the LLM team at Meta before he left to co-found Mistral AI.

On another note, the fact that AI startups are getting VCs to open their checkbooks is notable. The Googles, Apples, Metas, Amazons, Microsofts, etc. of the world will definitely make their plays in this space, but the fact that startups are being formed and are finding willing recruits is a signal that it’s not all game over just yet.

A mistral is a north wind that blows down Europe, and usually when it blows that is a sign of good weather in the days following. The startup may well be hoping for the same effect here.

San Francisco people often refer to their “uncanny valley,” but at the end of the day, Mensch said he believes that there has to be a credible effort from all around the world to build and take some ownership of how AI will develop. This is Mistral’s play to give France a shot at that.

Its plan is to use the funding to assemble what he describes as a “world class team” creating “the best open source models,” he said, and it will definitely now have some money to blow through to do just that.

More TechCrunch

As part of the update, Reddit also launched a dedicated AMA tab within the web post composer.

Reddit introduces new tools for ‘Ask Me Anything,’ its Q&A feature

Here are quick hits of the biggest news from the keynote as they are announced.

Google I/O 2024: Here’s everything Google just announced

LearnLM is already powering features across Google products, including in YouTube, Google’s Gemini apps, Google Search and Google Classroom.

LearnLM is Google’s new family of AI models for education

The official launch comes almost a year after YouTube began experimenting with AI-generated quizzes on its mobile app. 

Google is bringing AI-generated quizzes to academic videos on YouTube

Around 550 employees across autonomous vehicle company Motional have been laid off, according to information taken from WARN notice filings and sources at the company.  Earlier this week, TechCrunch reported…

Motional cut about 550 employees, around 40%, in recent restructuring, sources say

The keynote kicks off at 10 a.m. PT on Tuesday and will offer glimpses into the latest versions of Android, Wear OS and Android TV.

Google I/O 2024: Watch all of the AI, Android reveals

It ran 110 minutes, but Google managed to reference AI a whopping 121 times during Google I/O 2024 (by its own count). CEO Sundar Pichai referenced the figure to wrap…

Google mentioned ‘AI’ 120+ times during its I/O keynote

Google Play has a new discovery feature for apps, new ways to acquire users, updates to Play Points, and other enhancements to developer-facing tools.

Google Play preps a new full-screen app discovery feature and adds more developer tools

Soon, Android users will be able to drag and drop AI-generated images directly into their Gmail, Google Messages and other apps.

Gemini on Android becomes more capable and works with Gmail, Messages, YouTube and more

Veo can capture different visual and cinematic styles, including shots of landscapes and timelapses, and make edits and adjustments to already-generated footage.

Google Veo, a serious swing at AI-generated video, debuts at Google I/O 2024

In addition to the body of the emails themselves, the feature will also be able to analyze attachments, like PDFs.

Gemini comes to Gmail to summarize, draft emails, and more

The summaries are created based on Gemini’s analysis of insights from Google Maps’ community of more than 300 million contributors.

Google is bringing Gemini capabilities to Google Maps Platform

Google says that over 100,000 developers already tried the service.

Project IDX, Google’s next-gen IDE, is now in open beta

The system effectively listens for “conversation patterns commonly associated with scams” in-real time. 

Google will use Gemini to detect scams during calls

The standard Gemma models were only available in 2 billion and 7 billion parameter versions, making this quite a step up.

Google announces Gemma 2, a 27B-parameter version of its open model, launching in June

This is a great example of a company using generative AI to open its software to more users.

Google TalkBack will use Gemini to describe images for blind people

Firebase Genkit is an open source framework that enables developers to quickly build AI into new and existing applications.

Google launches Firebase Genkit, a new open source framework for building AI-powered apps

This will enable developers to use the on-device model to power their own AI features.

Google is building its Gemini Nano AI model into Chrome on the desktop

Google’s Circle to Search feature will now be able to solve more complex problems across psychics and math word problems. 

Circle to Search is now a better homework helper

People can now search using a video they upload combined with a text query to get an AI overview of the answers they need.

Google experiments with using video to search, thanks to Gemini AI

A search results page based on generative AI as its ranking mechanism will have wide-reaching consequences for online publishers.

Google will soon start using GenAI to organize some search results pages

Google has built a custom Gemini model for search to combine real-time information, Google’s ranking, long context and multimodal features.

Google is adding more AI to its search results

At its Google I/O developer conference, Google on Tuesday announced the next generation of its Tensor Processing Units (TPU) AI chips.

Google’s next-gen TPUs promise a 4.7x performance boost

Google is upgrading Gemini, its AI-powered chatbot, with features aimed at making the experience more ambient and contextually useful.

Google’s Gemini updates: How Project Astra is powering some of I/O’s big reveals

Veo can generate few-seconds-long 1080p video clips given a text prompt.

Google’s image-generating AI gets an upgrade

At Google I/O, Google announced upgrades to Gemini 1.5 Pro, including a bigger context window. .

Google’s generative AI can now analyze hours of video

The AI upgrade will make finding the right content more intuitive and less of a manual search process.

Google Photos introduces an AI search feature, Ask Photos

Apple released new data about anti-fraud measures related to its operation of the iOS App Store on Tuesday morning, trumpeting a claim that it stopped over $7 billion in “potentially…

Apple touts stopping $1.8B in App Store fraud last year in latest pitch to developers

Online travel agency Expedia is testing an AI assistant that bolsters features like search, itinerary building, trip planning, and real-time travel updates.

Expedia starts testing AI-powered features for search and travel planning

Welcome to TechCrunch Fintech! This week, we look at the drama around TabaPay deciding to not buy Synapse’s assets, as well as stocks dropping for a couple of fintechs, Monzo raising…

Inside TabaPay’s drama-filled decision to abandon its plans to buy Synapse’s assets