Notes from the AI Now Summit by Mistral | Koen van GilstSkip to main content>About>Lab>PhotographyNotes from the AI Now Summit by MistralMay 29, 20263 min readaiarticleI was in Paris the last few days to visit the AI Now Summit by Mistral AI, hoping to learn more about their models, plans for the future of European AI and more. My personal insights:
Mistral is no longer just a model company. They're building the full AI stack: compute, models, platforms & consultancy. They own the compute (a 40MW data center in Paris, more data centers coming soon, including one in Sweden). They focus on efficient, open and custom models that you own and can run on-prem. That seems to be their unique selling point compared to Anthropic or OpenAI.

The messaging was all about partnerships: collaborations with ASML, BNP Paribas, Amazon's Alexa+ and how they were helping them with AI to solve real problems. It was less about upcoming new models and tech innovation. Something I found disappointing. They did launch Vibe for Work, a product similar to Claude for Work.
When it comes to agentic, the harness is everything. In a talk by Pieter Stock he mentioned that the model alone isn't enough. With a harness you add context, persistence and learning. Reasoning is essential for this; it's what lets a system backtrack, recover from errors and stay transparent. Skills are the way for organisations to capture best practices, you develop these in cooperating with the AI agent.

Specialized small models are their strategy. Mistral showed several examples where small, fast and focused models outperform the big general-purpose ones when it comes to energy efficiency and speed: Document AI for OCR (used by the EU Patent Office to do large scale OCR), Voxtral for multilingual voice (powering Amazon's Alexa+ in Europe), and Robostral for industrial robotics with ASML. And also in token-heavy agentic applications, speed and efficiency are becoming as important as raw capability.
Sovereignty and on-prem are their selling points. BNP Paribas runs Mistral models on-prem for KYC in Belgium, with sensitive data staying within the bank's walls. Abanca is using agent orchestration to handle sensitive customer information at a huge scale (2 million customers in their app). For European companies in regulated industries, this is a good alternative to relying on US hyperscalers.
A talk that was a bit out of the ordinary and that I really enjoyed was about ancient papyrus documents: a research team from the Austrian Academy of Sciences finetuned a coding LLM by Mistral (Codestral) to read tiny snippets of millennia-old discarded papyri that had sat unpublished for decades. This work helps make a collection of 180,000 documents found in the Egyptian desert accessible, a job that would have taken more than 2000 years without AI. A beautiful example of how AI can also help the humanities.

All in all, the summit left me with a better picture of Mistral's vision for Europe an AI: maybe not to win the race for AGI (Artificial General Intelligence), but to become the European full-stack AI partner that delivers real return on investment NOW. Whether that pays off will depend on more European companies committing to this, but the combination of open models, on-prem deployment and enterprise partnerships could be appealing to many big organizations in the EU. And honestly, it's good to see a serious European player at the table. The days of blindly relying on US tech giants is coming to an end.Edit on GitHub© 2026 Koen van Gilstblueskymastodongithublinkedinv. 9.0.0 | d82a973

Mistral AI is presented as an entity building the entire AI stack, encompassing compute, models, platforms, and consultancy, distinguishing itself through ownership of infrastructure, specifically a 40-megawatt data center in Paris with plans for further expansion, including a center in Sweden. Their core competitive advantage lies in developing efficient, open, and custom models that can be run on-premise, which contrasts with the models offered by competitors such as Anthropic or OpenAI. The company emphasizes partnerships, collaborating with entities like ASML, BNP Paribas, and Amazon's Alexa+ to apply AI solutions to solve tangible problems, suggesting a focus on practical application over mere technological innovation.

A significant part of Mistral's strategy revolves around agentic systems, where the harnessing mechanism is identified as critical. As noted by Pieter Stock, relying solely on a model is insufficient; true utility requires adding context, persistence, and learning capabilities, alongside reasoning to enable systems to backtrack, recover from errors, and maintain transparency. Organizations can capture best practices by developing skills through cooperation with these AI agents.

The company pursues a strategy focused on specialized small models, demonstrating that these models can outperform large, general-purpose ones in terms of energy efficiency and speed. Specific examples cited include Document AI for Optical Character Recognition used by the EU Patent Office, Voxtral for multilingual voice applications powering Amazon's Alexa+ in Europe, and Robostral for industrial robotics collaboration with ASML. This highlights the growing importance of speed and efficiency alongside raw capability in token-heavy agentic applications.

Mistral strongly champions data sovereignty and on-premise deployment as key selling points, particularly for European companies in regulated sectors. This approach allows sensitive data to remain within organizational boundaries, as demonstrated by BNP Paribas running Mistral models on-premise for Know Your Customer (KYC) processes in Belgium, and Abanca utilizing agent orchestration to manage sensitive customer information across its application. This provides a viable alternative for European enterprises seeking to avoid reliance on US hyperscalers.

Beyond enterprise applications, the company engaged in research demonstrating AI’s capability in the humanities through an interesting project. A research team from the Austrian Academy of Sciences fine-tuned a coding LLM, Codestral, to process small snippets from ancient papyri, successfully making a collection of 180,000 documents found in the Egyptian desert accessible. This illustrates how advanced AI can contribute to historical and scholarly endeavors.

Ultimately, the summit suggested that Mistral’s vision for Europe is not necessarily about winning the race for Artificial General Intelligence but about establishing itself as the European full-stack AI partner capable of delivering immediate return on investment. This vision hinges on the combination of open models, on-prem deployment capabilities, and strong enterprise partnerships, which may appeal to large organizations within the EU. The overall message is that the era of blindly depending on US technology giants is concluding, positioning Mistral as a serious and appealing European player in the AI landscape.