In this study, we assessed over 80 countries across the Middle East, Central & Eastern Europe, Western Europe, Asia Pacific (APAC excluding China), Central Asia, Africa, North America (Canada, excluding US), and CALA. The analysis focuses on sovereign AI LLM development outside the US and China to assess how other nations are competing to build domestic LLM capabilities. Among these regions, over 55 active Sovereign LLM nations were identified for evaluation, where more than 170 models (latest active versions) were analyzed to determine ecosystem maturity in terms of Sovereign LLM ownership, foundational build, local language depth, and application.

The Sovereign AI Index is one of the foundational aspects of Counterpoint Research’s AI 360 Practice and is built on the success of our Global AI Cities Index. Our current research deep dives and maps different economies competing to build the best Sovereign AI models, infrastructure and services that can benefit not just their domestic landscape but also provide a template for other countries.

Commenting on the importance of Sovereign AI LLMs, Research Director Marc Einstein said, "The Sovereign AI LLM landscape has fundamentally shifted to national strategic consideration. Countries are prioritizing LLM Sovereignty as nations can introduce data security risks, alongside challenges in linguistic diversity, and tech independence. In this regard, with the rising political risk in the usage of foreign-developed LLMs, following the recent ban on Claude’s Fable 5, the case for Sovereign models further strengthens.”

Counterpoint Research’s Sovereign AI LLM Index

*The Sovereign AI LLM Spectrum measures how highly sovereign each LLM is, across dimensions with different strength patterns.

Sovereign AI LLMs at the leading edge of the Sovereign AI spectrum show consistent strength across the key evaluation dimensions: ownership, foundational model build, native-language depth, and real-world application. These capabilities explain their placement at the high end of the spectrum.

Middle East’s Sovereign AI LLM leadership:

• The Middle East emerged as the world’s top Sovereign AI LLM region, led by the UAE and Saudi Arabia.
• Four flagship models, Falcon H1, ALLaM, Jais 2, and K2 Think V2, sit at the leading edge of our Sovereign AI LLM spectrum.
• Their strengths include government-affiliated ownership, full Arabic dialect coverage, and broad citizen reach, for example UAE’s TAMM super app offering Sovereign AI native services.

Other leading Sovereign AI LLMs:

• Russia: GigaChat 3.1 Ultra is the UAE’s closest counterpart. Its competitive edge comes from vertical integration: Sber AI owns the full stack, including the 702B-parameter model, training on Sber’s Christofari supercomputer, and deployment through multiple apps, led by GigaChat.
• India: Sarvam 105B, developed by Sarvam AI under the IndiaAI Mission, marks India’s first foundational frontier model. It supports 22 Indian languages and is deployed through the Indus app, positioning India to move its large user base toward sovereign LLM adoption.
• Switzerland: The country secured a spot in the leading spectrum through consortium-driven public ownership.
• South Korea: A.X 3.1 stands out for the extensive reach of its consumer app, A. (A-Dot), supported by government backing.
• Japan and France: Japan’s PLaMo 3.0 Prime by Preferred Networks and France’s consortium-led Minerva, developed in collaboration with CINECA, both benefit from chatbot deployment. Their strengths stem from government backing and PPP-led ownership, respectively.

Analyzing the spectrum, Einstein added, “Positioned by capability and adoption, all 10 LLMs at the leading edge of the sovereignty spectrum are foundational builds, publicly funded or jointly developed by government and private partners, most optimized for deep native language coverage over English benchmarks. Across geographic representation, Technology Innovation Institute (Abu Dhabi, UAE) becomes the top government developer. Its Falcon H1 ranks No.1, with leadership across all dimensions.”

Citing key findings from the broader analysis, Einstein commented, “With a 56% share, adapted models built on a pre-existing base model dominate the overall Sovereign AI LLM landscape, outpacing foundational models developed from scratch. However, we observe countries taking a strategic, decisive step to move towards reducing reliance on open-source foreign bases. Looking at the developer’s side from a domain perspective, Sovereign LLMs are largely powered by AI companies and startups (Mistral AI, Sakana AI, Cohere, Sarvam AI) contributing to foundational and adapted models in their portfolio.”

South Korea has the largest Sovereign AI long-term investment at $1 trillion, while Japan benefits from METI’s GENIAC distributed industrial program across its domestic LLM developers. Consequently, both nations have gained notable momentum in the wider Sovereign AI LLM landscape.

Regional Landscape Summary:

Middle East: The Middle East emerged as the Sovereign AI flag bearer, driven by LLM Falcon H1 (UAE). Technology Innovation Institute (TII), the developer of Falcon H1, is currently the UAE’s top Sovereign AI LLM developer, followed by MBZUAI. Meanwhile, Iran is purely riding on adapted builds. The Middle East’s Sovereign LLM landscape is widening further with Qatar’s Fanar 3.0 confirmed to launch in December this year, Iraq’s upcoming govt-led Sovereign LLM, and Oman’s planned scaling of its national LLM Mu’een by MTCIT. Elsewhere, Kuwait, Jordan, and Bahrain are currently in the strategy phase.

Asia Pacific: South Korea, Japan, and India are APAC’s Sovereign AI LLM ecosystem leaders. India secured a place in the leading end of the spectrum with Sarvam 105B. India ranks higher on sovereignty due to its PPP structure. Japan/South Korea pay an "ownership penalty" despite richer ecosystems – South Korea accounts for the highest share of foundational LLMs at 17%, followed by Japan at 11%. Japan’s GENIAC-backed developers include Preferred Networks, NTT, Softbank, Rakuten, while South Korea’s MSIT-backed developers include LG AI research, SK Telecom, Naver- both ecosystems running largely on Govt-backed Private ownership.

• Central & Eastern Europe: Russia leads in scale and capability; the rest of the region relies heavily on adaptation with strong academic foundations. Russia dominates with 30% Sovereign LLMs share region-wide, including the region’s largest GigaChat 3.1 Ultra at 702B parameters. Turkey has emerged as a contender within the region with a 16% share of Sovereign models.
• Western Europe: Western Europe has shown strong sovereign AI capability with 59% foundational models regionally, led by France's Mistral AI. The region is mainly led by public ownership, riding on AI consortiums, along with academic and research sectors.
• Central Asia & Africa: Both regions are collectively ~80% dependent on adapted builds. Egypt and Kazakhstan lead through government investment and consumer deployment. Central Asia has yet to establish a foundational LLM and hence, remains reliant on adapted models. Africa exhibits limited foundational LLMs with Egypt and South Africa being the only primary source.
• North America (Canada): Private AI company Cohere solely carries Canada’s Sovereign AI status with two foundational LLMs – Tiny Aya (3.35B, SLM) and Command A (111B, Frontier).
• Central & Latin America (CALA): CALA’s Sovereign AI landscape begins and ends with Chile and Brazil with most other countries in the region being out of scope. Both Chile and Brazil possess adapted models with US-based Llama.

About Counterpoint Research 

Counterpoint Research is a global market research firm specializing in products across the technology ecosystem. We advise a diverse range of clients – from smartphone OEMs to chipmakers and channel players to Big Tech – through our offices located in the world's major innovation hubs, manufacturing clusters and commercial centers. Our analyst team, led by seasoned experts, engages with stakeholders across the enterprise – from the C-suite to professionals in strategy, analyst relations (AR), market intelligence (MI), business intelligence (BI), product and marketing – to deliver services spanning market data, industry thought leadership and consulting. Our core areas of coverage include AI, Automotive, Consumer Electronics, Displays, eSIM, IoT, Location Platforms, Macroeconomics, Manufacturing, Networks and Infrastructure, Semiconductors, Smartphones and Wearables. Visit our Insights page to explore our publicly available market data, insights and thought leadership, and to understand our focus, meet our analysts and start a conversation.