Yandex is pushing forward in the fields of artificial intelligence and robotics. Central to this effort is YandexGPT, its proprietary generative AI platform, which now powers a range of services and supports business and research applications.
The company has restructured its operations in recent years, spinning off several international companies to help globalize its technologies. These include Nebius, Avride, Yango, and Toloka—all born from Yandex but now independently run and internationally focused.
The Role Of YandexGPT In Russia’s AI Strategy
YandexGPT is a neural language model developed by Yandex LLC, playing a key role in the company’s generative AI plans. Launched initially as Yalm 2.0 in 2023, it quickly evolved, with version 5.0 released in February 2025.
This model is used widely across Yandex’s services. It powers the Alice voice assistant, serves business clients through the Yandex Cloud API, and has been integrated into applications for education, medicine, and content creation.
YandexGPT’s language abilities are tuned explicitly to Russian-language contexts. This helps meet the country’s goals under its national AI strategy, which aims for global leadership in select AI areas by 2030.
International Spin-Offs Lead Global Expansion
Yandex’s 2024 corporate restructuring led to the creation of spin-offs operating outside of Russia:
- Nebius, based in Amsterdam, develops AI cloud infrastructure. It has partnerships with Nvidia and plans to invest $1 billion in European AI infrastructure by mid-2025.
- Avride, headquartered in Austin, continues Yandex’s autonomous vehicle work and has partnerships with Hyundai and Uber.
- Yango Tech, part of the Dubai-based Yango Group, offers robotics and B2B AI solutions, including warehouse and delivery automation.
- Toloka, under the Nebius Group, focuses on AI data annotation services and has secured strategic investment from Bezos Expeditions.
These companies maintain Yandex’s technological foundations while operating independently, enabling focused strategies and reduced geopolitical risk.
AI Research, Applications, And Open Source Contributions
Yandex continues to invest heavily in AI research through its Yandex Research unit. The team works on natural language processing, machine learning, and computer vision. Key projects include decentralized model inference, speculative decoding, and tools for model compression.
Yandex has also made notable open-source contributions. Tools like Catboost, used for handling categorical data in machine learning, and Perforator, a performance profiling tool, have found wide adoption globally.
Applied AI Across Sectors
Yandexgpt is used in various sectors beyond search and virtual assistants. In education, it assists with creating tests and lesson plans. In healthcare, it supports clinical document processing. For content creation, it powers apps like Shedevrum that generate full posts with text and images.
Yandex’s AI also plays a critical role in advertising and personalization. The company’s Yarnit platform leverages AI for campaign design and keyword optimization.
A Distributed But Connected AI Ecosystem
Despite being legally separate, the spin-offs still carry Yandex’s “technological DNA.” Many inherited teams, infrastructure, and intellectual property from the parent company. This structure allows each to specialize while maintaining a shared foundation of innovation.
By focusing both on large foundational models like Yandexgpt and practical tools like Catboost, Yandex is executing a two-track AI strategy. Domestically, it strengthens Russia’s AI sovereignty. Internationally, it supports innovation through its spin-offs.
