Russian businesses turn to Chinese AI as sanctions render Western models too expensive
Lower prices and easy access are forcing Russian enterprises to quickly turn to Chinese AI models, leaving Western-made products behind. Industry insiders said the payment woes during the sanctions had accelerated a movement already driven by cost and increasing model quality.
Chinese-made AI is becoming the de facto foreign choice for more and more Russian firms, with models like Qwen, DeepSeek, GLM and Kimi soaring in popularity across the country’s main cloud platforms. Chinese models are generally far cheaper, install easily on Russian infrastructure, and avoid the payment issues that Western APIs have.
«For many business tasks, the difference between a flagship model priced at a notional dollar and a Chinese alternative at two cents is either infinitesimal or immaterial,» one industry executive told business daily newspaper Kommersant.
Yandex AI Studio clients consumed 597 billion tokens in the first half of 2026, an 18-fold increase from the previous year. Yandex’s own models remained at the top of the list, accounting for 54% of usage, but Chinese Qwen and DeepSeek combined for approximately 35%, far exceeding OpenAI’s GPT-OSS models’ 7.5%.
According to MWS Cloud, a Russian supplier of cloud and AI-powered solutions, the use of Chinese LLMs in the first half of 2026 hit 400 billion tokens, 11 times the amount in all of 2025. Qwen alone accounted for 261.1 billion tokens, with GLM and Kimi behind.
Read more: Alibaba releases Qwen 3.8-Max AI model to challenge Fable 5 and GPT Sol
Less expensive models, less payment headaches
Russian executives say the change is becoming less ideological and more commercial. Chinese models have gotten good enough that they can do a lot of intricate corporate work yet still be a lot cheaper than top US systems.
«The difference in quality between a flagship model, which costs a notional dollar, and a Chinese model, which costs two cents, is invisible or irrelevant across the vast majority of practical tasks,» Hybrid chief executive Dmitry Cheklov told Kommersant. He stated that the difficulty of paying for Western APIs is boosting the trend in Russia.
There is also a plus for open models. Companies can operate them on their servers or through Russian cloud vendors, meaning corporate data stays in the nation instead of being sent to a foreign developer. This approach has been the main way Russian enterprises use Chinese models when sensitive data is involved, Selectel, a server provider, added.
Read more: Anthropic launches Claude Opus 5 with premium AI performance
From economics and politics to business, technology and culture, Kursiv Uzbekistan brings you key news and in-depth analysis from Uzbekistan and around the world. To stay up to date and get the latest stories in real time, follow our Telegram channel.
