At AI4ALL, we’re always looking for ways to make AI more inclusive and transparent. That’s why we’re excited about DataChef’s newly launched LLM DutchRank.ai, a platform designed to show how well AI actually “speaks” Dutch.
LLM DutchRank.ai, developed by DataChef, is a transparent, community-powered, Dutch-first benchmarking platform. Built as a non-commercial, public-good initiative, it measures how foundation models actually perform in Dutch.
While global large language models (LLMs) excel in English, they often receive far less training in Dutch, resulting in weaker performance. This leads to hallucinations, poor reasoning, and reduced cultural nuance in Dutch-language outputs. Even with widespread AI adoption, there has been no publicly available, transparent benchmark for evaluating real-world performance in Dutch.
So DataChef built one. And they need your help!
DataChef is inviting Dutch speakers to help improve the benchmark by writing their own prompts and voting on model responses. This way, you can contribute to creating a more accurate and transparent benchmark for developers, organizations, and policymakers working with AI in the Dutch language.
The platform works in a simple but powerful way:
This approach results in a living leaderboard that shows which models actually perform best in Dutch, based on real usage and transparent comparison.
The first batch of models includes OpenAI, Anthropic, DeepSeek AI, xAI, and Google, with more being added soon.
DataChef, founded in 2019 in the Netherlands, has distinguished itself as one of the few same-side AI transformation partners, combining strategic navigation, execution and capability building into a single integrated approach.
They’ve also been a keen advocate for more inclusive AI, a commitment they backed by sponsoring AI4ALL’s inaugural AI4Her awards.
Whilst presenting the AI4Her Legacy award at the AI4All Gala, CEO and founder Ashkan Roshanayi explained DataChef’s commitment to responsible AI:
“Inclusivity is not just a concept. It is something that you live by. And that’s why I'm proud to be part of the AI4ALL community. It's not just about AI, it's about curiosity, bravery, and how these women are paving the path for future leaders.”
Through LLM DutchRank.ai, they are extending that same philosophy to the broader AI ecosystem, creating transparency, inclusivity, and opportunities for empowering more people to use the technology and benefit from AI.
DutchRank.ai reminds us that building responsible AI is a shared effort and gives the Dutch-speaking community a voice in how these systems evolve. It shows the future of responsible AI won’t be shaped by technology alone, but by the people who actively engage with it. With our prompts and feedback, we can all play a role in shaping how AI systems understand and respond to the Dutch language.
💫 Try it yourself: Type a prompt at DutchRank.ai. Compare answers. Vote.