At SenseTime's technology exchange day on April 10th, the company unveiled its latest multi-modal large model, "SenseNova V6," and the "SenseCore 2.0" system. This new version aims to integrate text, images, and videos, providing users with a more natural and richer interactive experience.

The SenseNova V6 series includes four versions. The most notable is SenseNova V6Pro, which boasts a 620 billion-parameter hybrid expert architecture, showcasing powerful multi-modal fusion capabilities. SenseNova V6Reasoner Pro enhances multi-modal reasoning capabilities, enabling deeper logical analysis. SenseNova V6Video focuses on video understanding, summarizing and deeply analyzing video content. SenseNova V6Omni is a lightweight, full-modal interactive model combining language, speech, and video for real-time interaction.

Demonstrations showcased SenseNova V6's unique multi-modal capabilities. Users could interact with the model using photos of handwritten math problems; the model not only solved them but also analyzed user answers, guiding users through the solution process via voice, even providing real-time assistance. This makes SenseNova V6 feel like a personal tutor.

SenseTime Technology

SenseTime co-founder, Linda Hua, stated that future interactions will inevitably be multi-modal, and SenseTime aims to master core technologies for these interactions. He noted the relative scarcity of domestic companies developing multi-modal reasoning and interaction capabilities, and SenseTime hopes to leverage its advantages in computer vision to preemptively establish a foothold in the multi-modal large model market.

Furthermore, SenseNova V6Pro's multi-modal capabilities are comparable to leading international models like Gemini 2.0Pro and GPT-4.5. SenseTime emphasizes strong reasoning, strong interaction, and long-term memory as three key technological breakthroughs. These capabilities allow the model to better understand human intent and foster more engaging user interactions.

SenseTime plans to integrate SenseNova V6 into real-world applications across various fields, including education, translation, and tourism. Collaborating with embodied AI company Fourier, SenseTime aims to equip robots with enhanced environmental understanding and human-robot interaction capabilities, truly realizing a more intelligent future.