Meta Platforms Inc. (NASDAQ:META) has emerged as the leading foundation for sovereign artificial intelligence models built on foreign open-source technology, according to new research from Counterpoint Research.
Meta’s Llama Leads Sovereign AI Models
Counterpoint found that 56% of the sovereign large language models it tracks are adapted models, meaning they are fine-tuned versions of existing open-source models. The remaining 44% are foundational models trained on native data.
Meta’s Llama accounts for 38% of adapted sovereign LLMs, making it the most widely used foreign base-model family. Alibaba Group Holding Ltd. (NYSE:BABA)-backed Qwen follows with a 20% share, while France’s Mistral holds 19%.
Counterpoint evaluated more than 170 models across more than 55 countries as part of its H1 2026 Sovereign AI LLM Index. The firm said Llama was the dominant base model across every region covered by the research.
See More: Top Value Stocks
Analyst Sees Push Toward Greater AI Autonomy
Counterpoint Research Director Marc Einstein said sovereign AI ultimately centers on control and strategic autonomy, but that independence exists on a spectrum.
“A slight majority of ‘sovereign’ LLMs are fine-tuned on open-source foreign bases, so true autonomy is still the exception,” Einstein said. He added that countries are increasingly moving toward foundational models built from scratch using domestic data.
Counterpoint said Qwen is gaining adoption across Central and Eastern Europe, Asia-Pacific and Africa, helped by its multilingual capabilities. Mistral, meanwhile, has gained traction in Europe, Asia-Pacific and the Middle East through partnerships including Abu Dhabi-based G42.
The research also points to growing use of other foundations, including Alphabet Inc.’s (NASDAQ:GOOGL) (NASDAQ:GOOG) Gemma, NVIDIA Corp. (NASDAQ:NVDA) Nemotron and DeepSeek, as countries diversify their AI technology stacks and reduce dependence on a single provider.
Photo via Shutterstock
Login to comment