Red River Insights - Open Source AI top 10
European open source AI is on fire!
Red River Insights - February 2025
Dear Friends,
“In my history, open source has always ultimately won.”
— Eric Schmidt, former CEO of Google
A few weeks ago, as the AI Summit turned the attention to Paris, a lot was said about the state of AI, and the spotlight was firmly on Europe’s strengths in open source AI. At the same time, the recent buzz around DeepSeek showed that open source is a way to innovate faster and in a more cost effective way than traditional closed source companies.
It’s no longer a secret to those who know us well: we just love open source. Open Source has historically given birth to fantastic companies around databases and DevTools, but it’s now proving equally exciting for AI, driving innovation and creating transformative opportunities in the field.
With recent breakthroughs in open source AI models, it’s only natural that our passion and curiosity for this space are growing exponentially. The rise of models like DeepSeek—powerful, cost-efficient, and freely available—proves that open source AI isn’t just catching up; it’s shaping the future of technology. That said, it’s important to acknowledge that for many in the community, the definition of open source is stricter—if the dataset used for training isn’t public, can we truly call it open source? This question remains at the heart of the discussion.
But open source AI is more than just accessible code—it’s a global movement driven by collaboration, transparency, and innovation. From foundational large language models to frameworks like TensorFlow and PyTorch, and platforms like Hugging Face that bring communities together, this ecosystem is rewriting the rules of AI development.
A key element of Open Source is indeed the power of the community around a project, its size, engagement, loyalty etc… So this month, we’re especially excited to introduce a new concept we worked on: the community score. The general ranking is - as usual - based on the momentum score but the community score provides complementary insights. This score reflects both the size and depth of engagement within an open-source community. We’ve had a lot of fun developing this framework, defining different layers of community involvement from core contributors to everyday supporters. We’ll share more about the science behind this score in a dedicated article very soon. Stay tuned!
In this edition, we present our Top 10 Open Source AI Companies in Europe—trailblazers who are not only advancing AI technology but also building the communities that will define its future. We’re excited to share their stories and explore how open source is driving the next wave of AI innovation.
RAMP's Open Source AI top10
We hope this sparks interesting conversations. If you have any comments or would like to suggest a startup that should be included, feel free to reach out to us. Abel, Chloé, Joseph and Olivier will be delighted to discuss these trends and rankings.
(Ranking established on 23/2/2025)
Europe's open source AI ecosystem is thriving, with companies pushing the boundaries of AI while fostering vibrant communities that drive innovation. This month's Top 10 list, curated through our proprietary RAMP platform, highlights the stars shaping the future of AI in Europe. Recent industry milestones underscore Europe's potential to lead the open source AI revolution, aligning with the belief that open source is key to technological sovereignty.
Here are the key trends defining this transformative landscape:
Foundational AI Models: Building the Bedrock of Intelligence
At the heart of the AI revolution are large language models (LLMs) and platforms that democratize access to AI. Mistral AI exemplifies this by providing developers and businesses with open and portable generative AI solutions. Their recent launch of Le Chat, a high-performance AI assistant, delivers responses at up to 1,000 words per second, making it the world’s fastest AI assistant and rivaling proprietary models at a fraction of the cost.
As data scientists worldwide collaborate to push the boundaries of AI, they need a place to build and share their breakthroughs. Hugging Face has become that hub, hosting over 500,000 models and 250,000 datasets. When DeepSeek released its latest model, it chose Hugging Face as the platform to distribute it—highlighting its role as the beating heart of open AI innovation.
Meanwhile, Black Forest Labs, a rising player from Germany, is making waves with its work in AI-generated imagery. Their Flux model has been turning heads for its ability to create stunning, high-quality visuals, offering a glimpse into the future of AI-powered creativity. This is yet another example of how open AI is driving innovation across fields beyond text, from art to design and beyond.
These companies exemplify Europe’s growing strength in foundational AI, reducing dependence on foreign tech giants and driving forward technological independence.
Neural Search and Conversational AI: Enhancing Human-Computer Interaction
The neural search sector is rapidly evolving, led by Jina AI and Text Cortex AI. Jina AI's latest release of Jina Embeddings 2.0 has significantly improved performance in large-scale data retrieval, empowering developers to build scalable search solutions within minutes. Meanwhile, Text Cortex AI has expanded its platform to support multilingual content generation, catering to global enterprises seeking personalized AI-driven copywriting. These advancements demonstrate how open source AI is making human-computer interactions more intuitive and accessible, addressing the growing demand for efficient data processing and user-centric solutions.
Speech and Language Processing: Bridging the Communication Gap
pyannoteAI has enhanced its speaker diarization technology, now enabling real-time speaker identification with improved accuracy—an innovation already being piloted in customer service and healthcare applications. Babelscape recently released a multilingual NLP toolkit that supports over 50 languages, positioning itself as a key player in cross-lingual AI solutions. These breakthroughs are pivotal in creating inclusive, globally connected AI systems that transcend linguistic barriers, showcasing Europe's ability to innovate in diverse AI domains.
Privacy-Preserving AI: Security Meets Open Source
With data privacy becoming paramount, Zama is leading the charge with its open source homomorphic encryption technology, enabling AI models to operate on encrypted data without ever exposing sensitive information. This innovation is gaining traction in industries like healthcare and finance, where privacy regulations are strict, allowing these sectors to deploy AI models and unlock their benefits—just as other industries have already done—without compromising security.
Zama's work aligns closely with Europe’s strong regulatory framework, such as GDPR, reinforcing data sovereignty and compliance. This trend underscores how open source solutions naturally fit within Europe’s commitment to transparency and accountability, offering a secure and scalable alternative to more opaque proprietary approaches.
MLOps and AI Infrastructure: Empowering Developers
ZenML and deepset are streamlining AI development through open source frameworks. ZenML's latest release introduces automated pipeline orchestration, reducing deployment time by 40%, while deepset's Haystack framework now includes enhanced document indexing and question-answering capabilities. By simplifying the development lifecycle, these companies are addressing the need for robust MLOps frameworks, enabling developers to deploy and manage AI solutions more efficiently and accelerating adoption across industries.
Emerging Trends and Insights from RAMP
Germany and France Lead the Way: Nine of the Top 10 companies are based in Germany and France, underscoring these countries' dominance in open source AI. Germany excels in neural search and MLOps, while France leads in LLMs and privacy-preserving AI.
Companies with high community engagement, such as Hugging Face, Mistral AI, and Zama, show that fostering active ecosystems accelerates innovation and adoption. The collective contributions of open source communities drive faster, more cost-effective advancements, enabling these companies to stay ahead of market expectations. By staying closely connected to their users and contributors, they can rapidly iterate, address real-world needs, and shape the future of AI in ways that proprietary models often struggle to match.
Privacy and Sovereignty in Focus: Zama's advancements in encryption exemplify how open source solutions align with Europe's regulatory environment, offering secure alternatives to proprietary models and enhancing data protection.
Conclusion
Europe's vibrant open source ecosystem, coupled with recent breakthroughs from companies like Mistral AI and Hugging Face, is reshaping the global AI landscape. By leveraging collaboration, transparency, and regulatory alignment, Europe is not only catching up but poised to lead the next chapter of AI innovation. As we continue to track these trends through RAMP, it's evident that open source is more than a development model—it's a strategic imperative for Europe's technological future.
More on RAMP's scoring method
The ranking of these startups is based on the estimated momentum of the company but the algorithm does not assess the quality or reliability of the products/solutions developed by these companies!
Find out about the algorithm behind this ranking and the way scores are calculated here: Cheat sheet on RAMP
In case you missed them, our latest top 10s are here:
- Janvier 25: Agentic AI
- December 24: France
- November 24: E-Commerce
- October 24: Global Defence
- September 24: Supply chain
- July 24: Travel Tech
- June 24: Developper tools
All the previous Top 10 are here.
In our first edition of Commit Pulse, our new newsletter dedicated to Open Source insights, Abel explored why betting on Open Source leads to greater sovereignty and how Europe is uniquely positioned to lead in a future where Open Source becomes the backbone of technological innovation. From regulatory alignment to a thriving developer ecosystem, Europe has the ingredients to shape the next wave of open innovation.
📖 Read the first edition here – and don’t forget to subscribe if you want to stay ahead of the latest trends, insights, and success stories shaping the Open Source landscape!
Read the first edition here
Portfolio news:
The Exploration Company achieved major milestones in February: it successfully completed a test campaign for its Typhoon rocket engine’s pre-burner (reaching ~16 seconds of stable combustion) and secured the German space agency DLR as an anchor customer for its Nyx capsule’s microgravity research services. (More here & here)
Resilience partnered with Europe’s top cancer center, Gustave Roussy, to leverage remote patient monitoring in early-stage drug trials. By deploying Resilience’s platform in first-in-human oncology studies, researchers can closely track treatment side effects in real time – a move aimed at accelerating clinical research and improving personalized cancer care. (More here)
Ada Health has been recognized as one of the Top 100 Digital Health and AI Companies to Watch in 2025 by The Medical Futurist Institute. This recognition highlights Ada’s pioneering role in AI-driven health assessments, which have already supported over 50 million symptom checks worldwide. As the company continues to expand its impact through strategic partnerships and clinical collaborations, its AI-powered platform is setting new standards for accessible and intelligent healthcare. (LinkedIn Post)
You received this newsletter because we met or worked together and we thought you might be interested in an update, if we got it wrong, feel free to unsubscribe with the link at the end!