SenseTime SenseNova 5.5: China’s first real-time multimodal AI model

SenseTime has unveiled SenseNova 5.5, an advanced version of its large language model (LLM) that includes SenseNova 5o—China’s first real-time multimodal model. SenseNova 5o marks a significant advancement in AI interaction, matching GPT-4o’s streaming features. This allows users to engage with the model in a conversational manner, making it ideal for real-time conversation and speech recognition applications.

Dr. Xu Li, Chairman and CEO of SenseTime, commented: “This is a pivotal year for large models as they transition from unimodal to multimodal. In response to user needs, SenseTime is enhancing interactivity. With applications driving model development and technological progress in multimodal streaming, we will witness unprecedented changes in human-AI interactions.” The upgraded SenseNova 5.5 offers a 30% performance improvement over its predecessor, SenseNova 5.0, released just two months earlier. Key enhancements include better mathematical reasoning, English proficiency, and command-following abilities.

To democratize access to advanced AI, SenseTime introduced a cost-effective edge-side large model, reducing the cost per device to as low as RMB 9.90 ($1.36) per year. This development is expected to accelerate adoption across various IoT devices. SenseTime also launched “Project $0 Go,” a free onboarding package for enterprises migrating from the OpenAI platform. This package includes 50 million tokens and API migration consulting services, aiming to lower entry barriers for businesses seeking to leverage SenseNova’s capabilities.

In support of edge-side AI, SenseTime released SenseChat Lite-5.5, featuring a 40% reduction in inference time to just 0.19 seconds and a 15% increase in inference speed, now reaching 90.2 words per second. Expanding its AI applications, SenseTime introduced Vimi, an AI avatar video generator that creates short clips with precise control over facial expressions and upper body movements from a single photo. This tool opens new possibilities in entertainment and interactive applications.

SenseTime also upgraded its Raccoon Series, a set of AI-native productivity tools. The Code Raccoon now boasts a five-fold improvement in response speed and a 10% increase in coding precision, while the Office Raccoon has expanded to include a consumer-facing webpage and a WeChat mini-app version. SenseTime’s large model technology is making significant impacts across various industries. In finance, it improves efficiency in compliance, marketing, and investment research. In agriculture, it reduces material use by 20% while increasing crop yields by 15%. In cultural tourism, it boosts travel planning and booking efficiency.

With over 3,000 government and corporate customers in technology, healthcare, finance, and programming sectors, SenseTime is solidifying its position as a leading AI player.