The Institute for Language and Speech Processing (ILSP) at the Athena Research Center – the only institute in Greece that has developed machines capable of answering questions and solving problems by mimicking human intelligence – has just launched its second large language model (LLM), Llama-Krikri.
Llama-Krikri follows in the footsteps of the institute’s first language model, Meltemi. While it may not rival the scale of DeepSeek or ChatGPT, it performs the same fundamental tasks. A conversation with the team behind it feels like a deep dive into the fascinating world of artificial neural networks, which expertly process and generate human language.
‘It’s like setting the language model to sift through your stored data and use it as “cheat sheet” to generate responses’
Why the name Llama-Krikri? “It’s a word association game,” explains Vassilis Katsouros, director of the ILSP at the Athena Research Center, with a smile. “The name comes from the base model we used – the open-source Llama 3.1 model from Meta – combined with the Greek equivalent of the llama, the kri-kri [a type of wild goat native to Crete].”
Llama-Krikri is significantly smaller in computational scale than DeepSeek, but by how much? “If DeepSeek’s neural network has 671 billion parameters, of which 37 billion are activated based on the user’s query, our model has only 8 billion parameters,” explains Katsouros. To put it simply, by translating the Greek language into numbers – since computers process numbers, not words – Llama-Krikri represents it using 8 billion numerical values. “It’s a different scale of investment,” he adds. “High-end language models require vast amounts of data, immense computational power, and significant financial resources.”
Digital aides
Llama-Krikri accomplishes everything that Meltemi did – and does a much better job. What exactly did Meltemi do? It empowered the Greek academic, research and business communities to build their own digital assistants. “While it doesn’t provide a service for answering direct queries from the general public, it does allow anyone with some programming knowledge to customize it to their needs, submit queries and receive answers based on their own data,” explains Sokratis Sofianopoulos, who, together with Prokopis Prokopidis, Georgios Paraskevopoulos, Leon Voukoutis and Dimitris Roussis – all scientific associates at ILSP – developed Llama-Krikri.
“It’s like setting the language model to sift through your stored data and use it as ‘cheat sheet’ to generate responses,” he adds. And these responses come in fluent modern Greek as well as in English. “Meltemi made a huge impact – tens of thousands of downloads on Hugging Face, a platform where AI developers share code. It was downloaded by students and academics, used in countless applications and research projects, boosted research activity in Greece and increased productivity in small and medium-sized enterprises. It offered an affordable solution for rapid project execution,” says Sofianopoulos. In addition, the institute has supported public organizations and businesses in building their own digital assistants. “Within the framework of the European Digital Innovation Hubs (EDIHs), we work with companies that request customized adaptations of our language model to suit their specific needs,” adds Katsouros.
So, what improvements does Llama-Krikri bring? “It has a much stronger memory – it doesn’t lose track during lengthy conversations. While the previous model could remember up to 8,000 words and sustain a dialogue for only a few minutes, Llama-Krikri can remember up to 128,000 words – an entire 250-page book – and can handle a sequence of multiple questions without forgetting what was said at the beginning. It can process larger volumes of text to generate its responses. It produces higher-quality text; it writes in the polytonic system, can compose in Ancient Greek, and translates into Modern Greek – something the previous model wasn’t trained to do. It also writes better code, as models aren’t limited to natural language – they also write computer languages. Additionally, it solves mathematical problems more accurately and exhibits stronger reasoning. In other words, it can perform complex tasks that require several intermediate steps of logical thinking,” explains Katsouros.
At the same time, the research team is creating and publishing a wealth of linguistic resources, including datasets that translate English language data into Greek. This effort allows for a thorough evaluation of Greek language model performance, and all of this is made openly available. As Maria Giagkou, a scientific associate at the institute, says, “Meltemi and Llama-Krikri are open-source models, meaning that any researcher or industry professional can freely examine and customize their workings. This is the power of open source. In contrast, OpenAI’s model is closed – you get what it gives you, with no possibility for tailoring to your specific needs.”
Katsouros further explains that the Greek ecosystem of small and medium-sized enterprises “needs technology in Greek that caters to its unique requirements and financial capabilities.” Moreover, as Sofianopoulos adds, a Greek company doesn’t need ChatGPT to support its services. It needs a smaller model – or perhaps even a second or third one – on which to assign specific tasks, all managed internally to avoid the risk of data leaks and large fees charged by OpenAI. This is especially important when services are provided via subscription, which is essential to ensure security. In such models, the cost is based on the number of word segments generated by each query and response. The more segments a word is broken into, the higher the cost. OpenAI and Anthropic often decompose each Greek word into multiple subwords – sometimes even character by character – thus generating Greek text by piecing these fragments together. “In our approach, we break down Greek words into the largest possible segments, significantly reducing the computational cost. On average, we split a Greek word into 1.5 segments, whereas other models break them into 2.5, 6, or even 7 segments,” he explains.
Cultural diet
But these are not the only reasons we need Greek language models. “Our language carries our culture and shapes the way we think and communicate in Greek. You need a service that, when you ask a culturally specific question, say, ‘Recommend a healthy breakfast,’ responds appropriately rather than suggesting something like ‘beans with bacon,’ which is a British concept of breakfast. It’s crucial to train large language models – and all language technologies – on texts originally written in Greek rather than on translations from other languages,” explains Giagkou. “It is also a matter of digital survival for our language. We know that unwritten languages can disappear; today, languages without a digital presence will also vanish. For a language to survive, it must have a digital footprint, which means being supported by language technologies, including large language models. And because we are a very small market, large companies show little commercial interest in our language. If we don’t invest in Greek language technologies, no one else will.”
A European DeepSeek?
Could Europe develop a DeepSeek? Katsouros responds: “It certainly could. DeepSeek utilized computational techniques that are well known in the research community. Similar methods have been adopted by the French company Mistral AI, one of Europe’s leading firms in language model development. What’s needed is more experimentation, continuous training, access to resources, hard work and constant adaptation to the latest technological advancements.
With open-source models, the effort is shared – the French build on the work of the Chinese, the Greeks build on the French, and so on – so a major European language model could indeed emerge and make a significant impact. However, Europe must remain actively involved and develop a more balanced regulatory framework for data copyrights – perhaps allowing 10% or 20% free use, similar to allowances made for educational purposes. [French President Emmanuel] Macron once observed that in Europe, we regulate too much and innovate too little.”
“Right now, a European alliance for language technologies has been formed, with 17 countries participating so far. The goal is to collect linguistic data and develop new large-scale language models. Additionally, two major European projects have been funded to create a multilingual model that will support all official European languages equally, in keeping with Europe’s multilingual policy,” says Giagkou.
Sofianopoulos is less optimistic, reminding us that Europe has not yet developed entirely new architectures for large language models – at least not on a scale that can compete with initiatives from the US and China. One initiative addressing part of this challenge is Pharos, “the Greek AI factory,” a project involving multiple domestic research and academic organizations. Key partners include the National Infrastructures for Research and Technology (GRNET), the Demokritos Center for Scientific Research, the National Technical University of Athens (NTUA), Athena RC and the Hellenic Corporation of Assets and Participations. Pharos will operate using computational resources from the high-performance computer (HPC) Daedalus.
“The more people that get involved, the faster we will see innovative results,” Katsouros concludes.