Mistral AI, the French company behind AI assistant Le Chat and several foundational models, is officially regarded as one of France’s most promising tech startups and is arguably the only European company that could compete with OpenAI. But compared to its $6 billion valuation, its global market share is still relatively low.  However, the recent launch of its chat assistant on mobile app stores was met with some hype, particularly in its home country. “Go and download Le Chat, which is made by Mistral, rather than ChatGPT by OpenAI — or something else,” French president Emmanuel Macron said in a TV interview ahead of the AI Action Summit in Paris. While this wave of attention may be encouraging, Mistral AI still faces challenges in competing with the likes of OpenAI — and in doing so while keeping up with its self-definition as “the world’s greenest and leading independent AI lab.” What is Mistral AI? Mistral AI has raised significant amounts of funding since its creation in 2023 with the ambition to “put frontier AI in the hands of everyone.” While this isn’t a direct jab at OpenAI, the slogan is meant to highlight the company’s advocacy for openness in AI. Its alternative to ChatGPT, chat assistant Le Chat, is now also available on iOS and Android. It reached 1 million downloads in the two weeks following its mobile release, even grabbing France’s top spot for free downloads on the iOS App Store. This comes in addition to Mistral AI’s suite of models, which includes:  Mistral Large 2, the primary large language model replacing Mistral Large. Pixtral Large, unveiled in 2024 as a new addition to the Pixtral family of multimodal models. Mistral Medium 3, released in May 2025 with the promise of providing efficiency without compromising performance, and best for coding and STEM tasks. Devstral, an AI model designed for coding and openly available under an Apache 2.0 license, meaning it can be used commercially without restriction. Codestral, an earlier generative AI model for code, but whose license banned commercial applications. “Les Ministraux,” a family of models optimized for edge devices such as phones. Mistral Saba, focused on Arabic language. In March 2025, the company introduced Mistral OCR, an optical character recognition (OCR) API that can turn any PDF into a text file to make it easier for AI models to ingest. Techcrunch event Join us at TechCrunch Sessions: AI Secure your spot for our leading AI industry event with speakers from OpenAI, Anthropic, and Cohere. For a limited time, tickets are just $292 for an entire day of expert talks, workshops, and potent networking. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 REGISTER NOW Who are Mistral AI’s founders? Mistral AI’s three founders share a background in AI research at major U.S. tech companies with significant operations in Paris. CEO Arthur Mensch used to work at Google’s DeepMind, while CTO Timothée Lacroix and chief scientist officer Guillaume Lample are former Meta staffers. Co-founding advisers also include Jean-Charles Samuelian-Werve (also a board member) and Charles Gorintin from health insurance startup Alan, as well as former digital minister Cédric O, which caused controversy due to his previous role. Are Mistral AI’s models open source? Not all of them. Mistral AI differentiates its premier models, whose weights are not available for commercial purposes, from its free models, for which it provides weight access under the Apache 2.0 license. Free models include research models such as Mistral NeMo, which was built in collaboration with Nvidia that the startup open-sourced in July 2024. How does Mistral AI make money? While many of Mistral AI’s offerings are free or now have free tiers, Mistral AI plans to drive some revenue from Le Chat’s paid tiers. Introduced in February 2025, Le Chat’s Pro plan is priced at $14.99 a month. On the purely B2B side, Mistral AI monetizes its premier models through APIs with usage-based pricing. Enterprises can also license these models, and the company likely also generates a significant share of its revenue from its strategic partnerships, some of which it highlighted during the Paris AI Summit. Overall, however, Mistral AI’s revenue is reportedly still in the eight-digit range, according to multiple sources. What partnerships has Mistral AI closed? In 2024, Mistral AI entered a deal with Microsoft that included a strategic partnership for distributing its AI models through Microsoft’s Azure platform and a €15 million investment. The U.K.’s Competition and Markets Authority (CMA) swiftly concluded that the deal didn’t qualify for investigation due to its small size. However, it also sparked some criticism in the EU.  In January 2025, Mistral AI signed a deal with press agency Agence France-Presse (AFP) to let Chat query the AFP’s entire text archive dating back to 1983. Mistral AI also secured strategic partnerships with France’s army and job agency, shipping giant CMA, German defense tech startup Helsing, IBM, Orange, and Stellantis. In May 2025, Mistral AI announced it would participate in the creation of an AI Campus in the Paris region, as part of a joint venture with UAE-investment firm MGX, NVIDIA, and France’s state-owned investment bank Bpifrance. How much funding has Mistral AI raised to date? As of February 2025, Mistral AI raised around €1 billion in capital to date, approximately $1.04 billion at the current exchange rate. This includes some debt financing, as well as several equity financing rounds raised in close succession. In June 2023, and before it even released its first models, Mistral AI raised a record $112 million seed round led by Lightspeed Venture Partners. Sources at the time said the seed round — Europe’s largest ever — valued the then-one-month-old startup at $260 million.  Other investors in this seed round included Bpifrance, Eric Schmidt, Exor Ventures, First Minute Capital, Headline, JCDecaux Holding, La Famiglia, LocalGlobe, Motier Ventures, Rodolphe Saadé, Sofina, and Xavier Niel. Only six months later, it closed a Series A of €385 million ($415 million at the time), at a reported valuation of $2 billion. The round was led by Andreessen Horowitz (a16z), with participation from existing backer Lightspeed, as well as BNP Paribas, CMA-CGM, Conviction, Elad Gil, General Catalyst, and Salesforce. The $16.3 million convertible investment that Microsoft made in Mistral AI as part of their partnership announced in February 2024 was presented as a Series A extension, implying an unchanged valuation. In June 2024, Mistral AI then raised €600 million in a mix of equity and debt (around $640 million at the exchange rate at the time). The long-rumored round was led by General Catalyst at a $6 billion valuation, with notable investors, including Cisco, IBM, Nvidia, Samsung Venture Investment Corporation, and others. What could a Mistral AI exit look like? Mistral is “not for sale,” Mensch said in January 2025 at the World Economic Forum in Davos. “Of course, [an IPO is] the plan.”  This makes sense, given how much the startup has raised so far: Even a large sale may not provide high enough multiples for its investors, not to mention sovereignty concerns depending on the acquirer.  However, the only way to definitely squash persistent acquisition rumors is to scale its revenue to levels that could even remotely justify its nearly $6 billion valuation. Either way, stay tuned. This story was originally published on February 28, 2025 and will be regularly updated.

The team behind Vercel’s V0, an AI-powered platform for web creation, has developed an AI model it claims excels at certain website development tasks. Available through an API, the model, called “v0-1.0-md,” can be prompted with text or images, and was “optimized for front-end and full-stack web development,” the Vercel team says. Currently in beta, it requires a V0 Premium plan ($20 per month) or Team plan ($30 per user per month) with usage-based billing enabled. The launch of V0’s model comes as more developers and companies look to adopt AI-powered tools for programming. According to a Stack Overflow survey last year, around 82% of developers reported that they’re using AI tools for writing code. Meanwhile, a quarter of startups in Y Combinator’s W25 batch have 95% of their codebases generated by AI, per YC managing partner Jared Friedman. Vercel’s model can “auto-fix” common coding issues, the Vercel team says, and it’s compatible with tools and SDKs that support OpenAI’s API format. Evaluated on web development frameworks like Next.js, the model can ingest up to 128,000 tokens in one go. Tokens are the raw bits of data that AI models work with, with a million tokens being equivalent to about 750,000 words (roughly 163,000 words longer than “War and Peace”). Vercel isn’t the only outfit developing tailored models for programming, it should be noted. Last month, JetBrains, the company behind a range of popular app development tools, debuted its first “open” AI coding model. Last week, Windsurf released a family of programming-focused models dubbed SWE-1. And just yesterday, Mistral unveiled a model, Devstral, tuned for particular developer tasks. Companies may be keen to develop — and embrace — AI-powered coding assistants, but models still struggle to produce quality software. Code-generating AI tends to introduce security vulnerabilities and errors, owing to weaknesses in areas like the ability to understand programming logic. Techcrunch event Join us at TechCrunch Sessions: AI Secure your spot for our leading AI industry event with speakers from OpenAI, Anthropic, and Cohere. For a limited time, tickets are just $292 for an entire day of expert talks, workshops, and potent networking. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 REGISTER NOW

Beyond performance and portability, its Apache 2.0 license offers a compelling proposition for commercial applications.Read More

AI startup Mistral on Wednesday announced a new AI model focused on coding: Devstral. Devstral, which Mistral says was developed in partnership with AI company All Hands AI, is openly available under an Apache 2.0 license, meaning it can be used commercially without restriction. Mistral claims that Devstral outperforms other open models like Google’s Gemma 3 27B and Chinese AI lab DeepSeek’s V3 on SWE-Bench Verified, a benchmark measuring coding skills. “Devstral excels at using tools to explore codebases, editing multiple files and power[ing] software engineering agents,” writes Mistral in a blog post provided to TechCrunch. “[I]t runs over code agent scaffolds such as OpenHands or SWE-Agent, which define the interface between the model and the test cases […] Devstral is light enough to run on a single [Nvidia] RTX 4090 or a Mac with 32GB RAM, making it an ideal choice for local deployment and on-device use.” Results from Mistral’s internal benchmarking evaluations of Devstral.Image Credits:Mistral Devstral arrives as AI coding assistants — and the models powering them — grow increasingly popular. Just last month, JetBrains, the company behind a range of popular app development tools, released its first “open” AI model for coding. In recent months, AI outfits including Google, Windsurf, and OpenAI have also unveiled models, both openly available and proprietary, optimized for programming tasks. AI models still struggle to code quality software — code-generating AI tends to introduce security vulnerabilities and errors, owing to weaknesses in areas like the ability to understand programming logic. Yet their promise to boost coding productivity is pushing companies — and developers — to rapidly adopt them. One recent poll found that 76% of devs used or were planning to use AI tools in their development processes last year. Mistral previously waded into the assistive programming space with Codestral, a generative model for code. But Codestral wasn’t released under a license that permitted devs to use the model for commercial applications; its license explicitly banned “any internal usage by employees in the context of [a] company’s business activities.” Devstral, which Mistral is calling a “research preview,” can be downloaded from AI development platforms including Hugging Face and also tapped through Mistral’s API. It’s priced at $0.1 per million input tokens and $0.3 per million output tokens, tokens being the raw bits of data that AI models work with. (A million tokens is equivalent to about 750,000 words, or roughly 163,000 words longer than “War and Peace.”) Techcrunch event Join us at TechCrunch Sessions: AI Secure your spot for our leading AI industry event with speakers from OpenAI, Anthropic, and Cohere. For a limited time, tickets are just $292 for an entire day of expert talks, workshops, and potent networking. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 REGISTER NOW Mistral says it’s “hard at work building a larger agentic coding model that will be available in the coming weeks.” Devstral isn’t a small model per se, but it’s on the smaller side at 24 billion parameters. (Parameters roughly correspond to a model’s problem-solving skills, and models with more parameters generally perform better than those with fewer parameters.) Mistral, founded in 2023, is a frontier model lab, aiming to build a range of AI-powered services, including a chatbot platform, Le Chat, and mobile apps. It’s backed by VCs including General Catalyst, and has raised over €1.1 billion (roughly $1.24 billion) to date. Mistral’s customers include BNP Paribas, AXA, and Mirakl. Devstral is Mistral’s third product launch this month. A few weeks ago, Mistral launched Mistral Medium 3, an efficient general-purpose model. Around the same time, the company rolled out Le Chat Enterprise, a corporate-focused chatbot service that offers tools like an AI “agent” builder and integrates Mistral’s models with third-party services like Gmail, Google Drive, and SharePoint.