Investment in artificial intelligence is experiencing unprecedented growth across the globe. European AI startups have secured around 55% more year-on-year investment in Q1 2025, according to data from Dealroom. Additionally, the European Commission recently also increased its investment efforts towards AI.
This sharp rise comes even as overall tech investment has slowed elsewhere, highlighting Europe’s growing momentum in the field. Nearly half of all unicorns are now AI-driven, underscoring just how central the sector has become to the next wave of innovation, as we highlighted in this article.
This surge in activity also reflects Europe’s ambition to carve out a leading role in the global AI race, challenging dominant players like the US and China. Yet Europe’s approach is distinctive. As we explored in “AI in the EU: A defining moment of uncertainty and opportunity“, the EU’s regulatory path aims to set standards for trustworthy and human-centric AI.
In connection with this, we’ve just compiled a list of 20 exceptionally promising AI startups founded between 2022 and 2025. The selection is presented in alphabetical order.
Adaptive ML: Headquartered in Paris with an office in New York, this startup is working on a method to improve the process known as reinforcement learning from human feedback, or RLHF. Their LLM platform enables continuous improvement of language models by learning from real-time user interactions, allowing businesses to personalise AI outputs and enhance decision-making efficiency. The platform also helps companies train and deploy language models using user feedback. Since its funding in 2023, Adaptative ML has successfully raised €19 million.
By simplifying complex processes like fine-tuning and preference optimisation, Adaptive ML reduces the time required to bring AI-powered products to market. Its technology supports a range of applications, from customer support to software development, and aims to make adaptive AI both accessible and scalable.
Bioptimus: Founded in 2024 and based in Paris, Bioptimus is developing a universal AI foundation model that integrates biological data across multiple scales, from molecules to whole organisms, to accelerate discoveries in biology and medicine. Backed by €67 million in funding and led by former Google DeepMind and Owkin scientists, the startup combines advanced computing infrastructure with deep expertise in life sciences to drive innovation in the field.
Bioptimus has released H-optimus-0, reportedly one of the largest open-source AI foundation models for pathology, trained on over 500,000 histopathology slides. This model achieves state-of-the-art performance in cancer detection and gene expression prediction. Looking ahead, Bioptimus is developing a new multi-scale, multi-modal model designed to simulate complex biological processes, with applications spanning biotechnology, medicine, and cosmetics.
Black Forest Labs: Based in Freiburg im Breisgau, Germany, Black Forest Labs is developing next-generation generative deep learning models for image and video creation. With a strong focus on accessibility and transparency, the startup aims to democratise powerful AI tools for content generation. Since its launch in 2024, Black Forest Labs has raised €28 million.
The startup’s suite includes state-of-the-art models such as FLUX 1.1 Pro for high-resolution text-to-image generation, alongside creative tools such as FLUX.1 Fill, Depth, and Canny for advanced customisation. Black Forest Labs’ technology integrates with platforms including Replicate, Together.ai, and Freepik. By combining cutting-edge generative models with a commitment to openness and usability, Black Forest Labs is helping to redefine the landscape of AI-powered media creation and sharing.
Cogna: The London-based SaaS platform helps businesses streamline their data integration processes across multiple domains. Founded in 2023, Cogna offers tailored, AI-powered software that identifies operational pain points and delivers custom solutions within weeks. So far, the startup has raised almost €19.3 million to accelerate its mission of transforming enterprise software development.
Cogna’s platform allows non-technical teams to describe their challenges in plain language, which the AI system then translates into fully functional software that integrates with existing tools like SAP, Oracle, Databricks, and Snowflake. With clients including Cadent Gas and Network Plus, the startup is already demonstrating how AI can reduce development costs and timelines in complex industries such as utilities, manufacturing, and construction, making enterprise-grade digital solutions faster and more accessible.
CuspAI: This AI startup is developing a platform for next-generation materials to address global sustainability and renewable energy concerns. The Cambridge-based startup, founded in 2024, uses generative AI, deep learning, and molecular simulation to accelerate material design. Like a search engine for materials, CuspAI allows users to specify the properties they need, and the platform generates corresponding molecular structures on demand.
By inverting the traditional approach to material discovery, starting with the desired outcome rather than trial-and-error, CuspAI is helping to fast-track innovations in areas like carbon capture, battery storage, and renewable infrastructure. The startup has raised €26.6 million from investors including Hoxton Ventures and Lightspeed Venture Partners. Its collaboration with Meta’s Fundamental AI Research (FAIR) team further highlights its position at the forefront of AI-driven materials science.
Emmi AI: This Austrian DeepTech startup is merging physics and AI to unlock a new era in industrial engineering. Emmi AI was founded in 2024 to tackle a major bottleneck in the sector: traditional simulations are slow and computationally expensive, often holding back innovation. With €15 million in funding, the team are building intelligent tools to speed up and simplify complex engineering processes.
Their approach starts with computational fluid dynamics and aims to expand into real-time, multi-physics simulations. By developing AI models that can replace heavy-duty simulations, Emmi AI helps companies in sectors like aviation, energy, and automotive to streamline design, planning, and testing. It’s a bold vision that blends scientific rigour with smart automation, making advanced engineering faster, more efficient, and more accessible.
Flower Labs: Tapping into the power of federated learning, this Hamburg-based startup is enabling companies to train AI models on distributed, sensitive data. By avoiding centralised data collection, the startup helps organisations like Nokia, Porsche, Brave, and Banking Circle leverage data from silos or user devices without compromising privacy.
Its open-source framework is designed to unlock new capabilities in AI by tapping into orders of magnitude more data than currently used. Flower’s solution aims to change the centralisation of AI data by making distributed training not only possible but also efficient and scalable. The startup’s mission is to expand access to meaningful AI innovation without sacrificing data security or control. It represents a growing movement towards privacy-preserving, decentralised AI infrastructure. Launched in 2023, the startup has recently obtained €20.7 million in financial backing.
Granola: The London-based startup launched in 2023, focuses on boosting productivity and organisation through a discreet, AI-powered notepad. Unlike traditional meeting bots, its tool transcribes and summarises meetings directly from the user’s device without the need to join calls, offering a seamless and non-intrusive experience. Designed for professionals juggling back-to-back meetings, it helps users capture key points and action items effortlessly.
The platform supports customised templates for different meeting types, such as sales calls, interviews, and team stand-ups, and automatically structures notes into clear summaries. With €58 million raised to date, Granola is expanding its offering with features that assist users beyond the meeting itself, such as drafting follow-up emails and generating content.
H Company: Founded in 2023 and based in Paris, this innovative startup creates AI models to boost worker productivity and make better decisions. Their focus is on advanced action models that automate tasks and support complex decision-making, making operations smoother for organisations. H Company’s goal is to develop highly advanced AI systems, with a long-term vision of achieving Artificial General Intelligence (AGI). Their latest product, Runner H, is a web agent that performs better than leading benchmarks, demonstrating its strength in real-world tasks. Since its start, H Company has raised €175.5 million in funding.
The startup has also developed what it claims is the best Vision Language Model (VLM) for understanding user interfaces. Its VLM topped the Screenspot benchmark by predicting user interactions (like mouse clicks) from simple instructions and screenshots. This makes H Company a leader in creating next-generation AI agents that can navigate and operate within digital environments, helping businesses work more efficiently and reduce manual tasks.
Latent Labs: Founded in 2023 and headquartered in London with offices in San Francisco, Latent Labs is an AI lab working to make biology programmable by developing foundational models for health and sustainability breakthroughs. The startup specialises in advanced generative AI for de novo protein design, enabling the creation of novel proteins not found in nature. These models power their platform, allowing researchers to computationally generate and optimise therapeutic molecules like antibodies and enzymes.
Operating on a hybrid model, Latent Labs provides drugmakers direct access to its AI platform and engages in collaborative partnerships. The startup has raised €44 million and focuses on targeting complex biological challenges, streamlining research through AI and in-house wet lab validation.
Legora: Based in Stockholm, Legora is a fast-growing LegalTech startup offering a collaborative AI platform designed to help legal professionals review documents, conduct research, and draft with greater efficiency. Its tools streamline complex workflows, transforming legal documents into interactive grids for faster analysis and integrating seamlessly into environments like Microsoft Word. Founded in 2024, Legora is rapidly expanding its global presence with offices in Stockholm, London, and New York, and has raised €102 million in funding.
Serving thousands of lawyers at top firms in over 20 countries, the innovative startup is redefining how legal professionals engage with technology. Its AI-powered features automate repetitive tasks, enabling legal teams to focus on high-value work. Their mission is to empower lawyers through intelligent collaboration and reshape the legal sector with advanced AI.
Lovable: Launched in 2023 in Stockholm, Lovable is an innovative startup making it easier for people to build software, even if they don’t have advanced programming skills. Its user-friendly platform allows professionals from different industries to turn their ideas into working applications quickly and easily.
By simplifying the development process and removing common technical barriers, Lovable helps bridge the gap between a concept and a finished product. Its tools are designed to fit real-world workflows, making software creation more accessible and efficient. So far, the startup has secured €18.5 million in funds.
Luzia: The Madrid-based AI startup, launched in 2023, aims to simplify and democratise AI. Its flagship product is a personal assistant that integrates with WhatsApp, enabling users to manage daily tasks through conversational interaction. Prioritising usability and privacy, Luzia offers features like audio transcription, text translation, and chatbot-driven content creation. The startup has secured nearly €40 million in funding, reflecting strong investor confidence in its scalable, user-focused AI productivity solutions.
Luzia’s assistant, powered by advanced NLP and machine learning, responds to text and voice inputs, adapting to individual user needs. It supports various applications, from educational tasks to professional communication and personal organisation. The AI continuously learns and evolves, serving as a versatile tool for enhancing daily life while ensuring secure, end-to-end encrypted interactions.
Mistral AI: Headquartered in Paris, this startup is building open and portable generative AI solutions for developers and businesses. Mistral AI, founded in 2023, focuses on creating state-of-the-art models for natural language processing and complex problem-solving. It has raised €1.1 billion to support its mission of making cutting-edge AI widely accessible. Established by former AI researchers from Google DeepMind and Meta, the startup has quickly become a standout in Europe’s tech ecosystem with its commitment to open-source development.
Mistral AI has released several high-performing language models, including Mistral 7B, Mixtral 8x7B, and Mistral Large 2, designed for diverse applications from enterprise use to integration in developer tools. With its focus on efficiency, transparency, and broad usability, Mistral AI is playing a central role in shaping the future of European AI innovation.
Nscale: Based in London, Nscale is a cloud-based platform that aims to build AI infrastructure and provide powerful computing tools for companies. Since launching in 2024, it has raised €163 million to support its goal of making AI development faster and easier. Nscale offers everything needed to build and run AI systems, from training and improving models to using them at scale. Its platform gives access to advanced computing power, both physical and virtual, as well as flexible services that don’t require managing servers. This makes it easier for AI teams and businesses to handle demanding projects efficiently.
Nscale offers high-performance GPU computing tools that streamline the entire AI development lifecycle, from training and fine-tuning to large-scale deployment. By providing seamless access to scalable computing power and AI frameworks like TensorFlow and PyTorch, Nscale removes the typical bottlenecks faced by teams building complex AI models.
Omnea: Operating from London, Omnea is an AI-powered procurement platform designed to help companies streamline the entire supplier lifecycle. The platform automates key processes such as vendor onboarding, intelligent intake, purchase order creation, risk management, and compliance tracking. By simplifying purchasing requests and automating approval workflows, Omnea enhances operational efficiency across multiple departments, making procurement faster, more transparent, and less error-prone.
Founded in 2022, Omnea has successfully raised almost €22 million to fuel the expansion of its end-to-end procurement capabilities and to establish a stronger presence in international markets. The startup addresses the complexity of traditional procurement by leveraging intelligent automation and comprehensive supplier management, empowering businesses to take control of their procurement processes with greater ease and security.
Orasio: Founded in Paris in 2025, the innovative startup specialises in developing AI solutions for real-time video analysis aimed at enhancing public safety and security. Their technology is designed to detect and interpret specific situations, such as identifying individuals carrying weapons or recognising the onset of fires, within live video feeds. This capability is particularly beneficial for internal security forces, municipalities, and private entities managing sensitive sites.
Due to the growing demand for smarter security solutions, Orasio has already landed €14 million in funding. Their technology is built to be flexible, allowing it to work smoothly across different systems, from cloud platforms to on-site servers and even directly on security cameras, while maintaining strong performance and privacy standards.
Prem Labs: Based in Lugano, Switzerland, Prem Labs is an applied research lab focused on open-source AI innovation. Founded in 2023, the lab is dedicated to advancing sovereign generative AI systems, with a mission to make advanced AI accessible, private, and fully customisable. Prem Labs has raised €24 million to fuel open innovation and support its vision of democratising AI technology.
Prem AI empowers both businesses and individuals to fine-tune and deploy their large language models without requiring deep machine learning expertise. With a strong emphasis on privacy and data sovereignty, its tools, such as the Autonomous Fine-Tuning Agent and the encrypted TrustML framework, enable secure, on-premises AI deployment. Open-source projects like ‘Prem-1B-SQL’ and ‘PremSQL’ further support the startup’s goal of making AI efficient and reliable, even on low-resource devices, helping ensure that powerful AI remains accessible to all.
Tessl: Founded in 2024 and based in London, Tessl is reshaping software development with an AI-native approach. Rather than simply using AI to support human developers, Tessl places AI at the core of the development process. Its platform generates secure, high-quality, and self-maintaining code tailored to each user’s needs, redefining how software is built in the AI era. By translating natural language specifications into modular, test-driven code, Tessl enables its platform to autonomously implement, test, and maintain applications, significantly reducing manual effort and accelerating development.
The applications of Tessl’s technology are especially relevant to enterprise software. It automates critical but routine tasks such as dependency updates, security patching, and API change management, freeing developers to focus on strategic initiatives. This enables faster release cycles, enhanced code quality, and broader access for non-experts. With €110 million in funding, Tessl is positioned to shape a more efficient, inclusive future for software development through deep AI integration.
Vespa.ai: Headquartered in Trondheim, Norway, the disruptive startup specialises in large-scale, real-time AI applications. Originating from Yahoo’s internal search technology, Vespa.ai was open-sourced in 2017 and became an independent company in 2023. The platform combines vector search, machine learning, and retrieval-augmented generation (RAG) to enable low-latency computations over extensive datasets, supporting industries such as e-commerce, financial services, and healthcare.
Since its spinout in 2023, Vespa.ai has raised €27.5 million in Series A funding led by Blossom Capital. Its clients include Spotify, Wix, and Yahoo, which use Vespa’s capabilities for tailored content delivery and advanced search functionalities.
By the way: If you’re a corporate or investor looking for exciting startups in a specific market for a potential investment or acquisition, check out our Startup Sourcing Service!
Read the orginal article: https://www.eu-startups.com/2025/05/the-ai-uprising-20-european-startups-rewriting-the-rules-in-2025/
