Windy Shao

windysjf@hotmail.com

AT the recent Global AI Device Expo 2025 in Shenzhen, SenseTime Technology stole the spotlight with its latest AI-powered chess-playing robots. The humanoid robot captured the attention of visitors young and old, offering not just a game of chess but a bridge across generations.

Artemey Melnikov, an 1800-rated chess player from Belarus, was among the first to test the robot. “This is my first time playing against a robot,” he shared with excitement. “This one is extremely entertaining — fast enough and very logical. It spotted my mistakes within seconds.”

AI chess robots

As smartphone and tablet use continues to rise, parents seek alternatives to reduce screen time for their children. The SenseTime AI chess robot offers a real, hands-on, screen-free interactive experience that engages both the young and old. It not only fosters family bonding but also helps bridge the digital divide for older users, making technology accessible and enjoyable.

Powered by SenseTime’s proprietary AI models, the robot uses a combination of advanced AI vision and robotic arm technologies to execute precise, millimeter-level chess moves. With certification from the Chess and Card Management Center of the State Sports General Administration and the Chinese Chess Association, the robot provides professional-level chess instruction, helping users improve and even reach official levels of play from the comfort of their homes.

Chen Jiafeng, a staff member of SenseTime’s AI chess robot division, shared insights during an interview with Shenzhen Daily.

“Our new product development is based on SenseTime’s own AI large models,” Chen explained. “These empower various industries, including culture and tourism, education, and beyond. We began three years ago with our very first Chinese chess robot. Today, we support chess, Go, and other board games. We are the only company globally that can mass-produce such robots.”

The robots have already found homes not only in China but also in Japan, South Korea, and parts of Europe. “The market potential is enormous,” Chen said. “We want our products to reach every corner of the world and every family that needs one.”

To achieve this, SenseTime has emphasized localization. “For overseas markets, we make product adjustments to better match local preferences, including colors and language localization,” Chen noted. “Europe is already a key focus for us, especially with our chess robot.”

Chen also introduced one of the robot’s most innovative features — a built-in English conversation companion tailored for children. “Based on their grade level, the robot talks with kids using age-appropriate vocabulary, facilitating daily conversations,” he said. “This not only helps with language learning but also encourages more family interaction. Many elderly family members still enjoy playing Chinese chess, so a single device can bring joy to multiple generations.”

A smart future

The chess robot is only one of the products from SenseTime, a leading AI software company providing AI technologies across industry verticals, such as generative AI, traditional AI, and smart auto.

At the forefront of AI development, SenseTime is driving progress toward a smarter, AI-empowered future through continuous innovation and technological breakthroughs.

The firm invests heavily in research that enables it to deliver industry-leading AI capabilities. Its expertise spans multimodal and multi-task models, with strong capabilities in perception intelligence, natural language processing, decision intelligence, AI-powered content generation, as well as AI chips, sensors, and computing infrastructure.

Excellent AI infrastructure

At the heart of this innovation lies SenseCore, SenseTime’s proprietary AI infrastructure that integrates computing power, algorithms, and platforms. It powers the development of the SenseNova foundational model suite and R&D system, enabling efficient, low-cost execution of general AI tasks.

The SenseNova Unified Large Model excels in both deep reasoning and multimodal information processing. It recently claimed first place on two comprehensive benchmarks — one testing language capabilities and the other visual content understanding.

In the 2024 SuperCLUE Chinese Large Model Benchmark Report, the SenseNova Unified Large Model ranked first in China with a score of 68.3, tying with DeepSeek V3. It also achieved the highest score in multimodal evaluations on OpenCompass, surpassing GPT-4o.

In SuperCLUE’s global rankings for 2024, SenseNova earned an outstanding score of 81.8 in the humanities, taking the top spot globally and outperforming OpenAI’s o1 model. It also took first place in China for scientific tasks, with a score of 78.2 in computational evaluation.

Thanks to its native multimodal capabilities, the model goes beyond traditional AI “reading” and “reasoning” to perform complex tasks such as deciphering unclear handwriting, extracting data from charts, and generating creative literary content.

SenseTime is also actively contributing to the formulation of national and international standards around data security, privacy protection, ethical AI, and sustainability.

The company collaborates with leading institutions to advance responsible AI development. With offices in Japan, Singapore, Saudi Arabia, the UAE, Malaysia, South Korea, and established presences in Germany, Thailand, Indonesia, and the Philippines, the company is expanding its global footprint and influence across diverse markets.